w. " ”T" r r ‘.’_- ‘. ' M ' fi- . -~1 ‘ ‘ If ‘ . ' _ ‘ . Ir * WW. ‘- *IW.‘ ["11 “to ‘ ifi V. ° ‘ ..}~ 4» .. . g .v I‘ A ‘f \’ . ,A ‘0 9w ‘ I I ' D t ' '.,., ' 4k . 'lf ' r. m ,4 ' { P ‘r' ‘ |'~. ’Ino‘uilfif 4-4 Jr"; . ‘ L ”29‘ We: -‘ ' m = 2.2:: km m um .14 n Y" ‘ . , _ i _ .61; v. 1&2. . :\ C’Qfa. “6+” ' .3” ' ‘ a. y I ' -' - 1‘ " ‘ '\:~ ,‘ 3' A 1-1. “5“ "u ‘ :L ‘1» .- .»_ ~";+}ann.:m-~i' ‘ - of: - . . . a.” er%.h. , .,¥....,7‘.?_1;:.‘ r; «1*: ‘ ‘1" ' x :3 1 ‘ ~~I9¢‘ .xkr'crmz' . - . ‘ ~ .- ' v’ ‘. N‘ A‘ " ’- v '-" . 9 1. ‘ ’1 ‘ ‘I I. '- .‘fifi II -I “7 p L‘ .3 .- w .v .» s. 145‘. an; 1. arm, .- 1r“) _ . . _ ‘ 5‘ :2: L.‘ g ' Ch L . 1‘ r.‘ J 2 A OJA‘ F}: :21 ' ’ i A on ‘G. .171". a 1 _{.' an “ 94-2. -I‘ .‘Ji. J '.| {-t'r '.'""'~'<" ‘ Ml,‘ _ ‘ .d‘ a." Q‘fi'Jeuur' LIBRARY MiCifi 5% Stave University 6’5 1?? kl .r". J ‘11 .53 This is to certify that the thesis entitled Carbon-13 Nuclear Magnetic Resonance Studies on the Solution Behavior of Sugar Phosphates (Part I) and Ribulose Bisphosphate Carboxylase/Oxygenase: Catalysis and Activation (Part 11) presented by John William Pierce, Jr. has been accepted towards fulfillment of the requirements for Ph. D. degree in _B_chhemj_s_try ‘21s- ZflZv-‘iw Major professor Date WWW" WM???» / 92], I? :5? 0—7 639 OVERDUE FINES: 25¢ per day per item ' ‘: (3::“4‘ L 1.1,. alj‘fi- . W: \ .‘ 1. _ _ ‘1. “mm A Y I 1 Place in book return to move charge from circulation records 15413 6 4 E2602 CARBON-13 NUCLEAR MAGNETIC RESONANCE STUDIES ON THE SOLUTION BEHANIOR OF SUGAR PHOSPHATES (PART I) AND RIBULOSE BISPHOSPHATE CARBOXYLASE/OXYGENASE: CATALYSIS AND ACTIVATION (PART II) By John William Pierce, Jr. A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Biochemistry 1980 95%) as assayed by 13C NMR using a short pulse width (10 22 us, 55°) and long delay time (10 s) to facilitate aldononitrile detec- tion. The racemic mixture of glyceronitrile phosphate was adjusted to pH 1.7 2 0.1 with 2 M HCl and hydrogenolyzed directly to DL-glyceraldehyde- 3-P without further purification. Epimeric tetrono- and pentononitrile phOSphates were purified by ion-exchange chromatography on Dowex 1-X8 (ZOO-400 mesh) in the formate form at 4°C by employing linear gradients of sodium formate (Table 2 and Figure 2A). The nitrile phosphate with .gi§42,3-hydroxyl groups is the major product and is eluted last under these conditions. Column capacity exceeds 6 mmol of aldononitrile phosphate. Yields of the glycero-, tetrono-, and pentononitrile phOSphates after cyanide condensation and chromatography are 85%. Fractions containing the aldononitrile phosphates were pooled and adjusted to pH 1.5 with Dowex 50(H+). After filtration, the acidic solutions were concentrated to 100 mL ig_vgggg_at 30°C and extracted continuously with diethyl ether overnight at 4°C to remove formic acid. The aqueous, acidic solutions were recovered, concentrated in vagug_at 30°C to approximately 10 mL, and adjusted to pH §_2 with Dowex 50(H+) prior to hydrogenolysis. Aldose Phosphates. Palladium-barium sulfate (5%, 62 mg per mmol of nitrile) was weighed into a side-arm flask, 5-10 mL of H20 was added, and the suspension was reduced with Hz for 15-20 min at atmospheric pressure and 25°C with efficient stirring. During this period, the catalyst changed from a brown to a whitish gray color. The solution of aldononitrile phosphate adjusted to pH 1.7 as described below was added from an addition funnel into the reduction vessel which was filled and evacuated three times prior to a final charging with H2. The 23 FIGURE Separation of DL-[1-13§]xylononitrile-S-P and DL-[I-1 Clexononitrile- 5- P and1 C NMR analyses of the products after hydrogenolysis over palladium. (A) Chromatography of the 2-epimeric pentononitrile phosphates on a 2.2 x 51 cm Dowex 1-X8 (ZOO-400 mesh) column in the formate form at 4°C developed with a linear gradient of sodium formate (3000 mL, 0. 05-0. 8 M, pH 3. 9). Column effluent was assayed for radioactivity and total phgsphate. The xylo epimer was eluted before the lyxo epimer. (B and C) C NMR Analyses showing resonances dgg to the enriched carbons )of the reduc- tion pfgducts from (B) DL-[I C]xylononitrile-5-P and (C) DL-[l Clexononitrile-S- P; C- I resonances of the 1- 3C- -en- riched a- and a-furanose ang hydrated forms of DL-aldose-S- P appear at approximately 100 ppm; [1-1 CJ- I-amino-I -deoxyalditol- -5- P (a) appears at approximately 43 ppm; and resonances due to natural abun- dance C of acetic acid (Ac), used to adjust the pH prior to hydrogenolysis, appear at approximately 23 and 180 ppm. Spectra were obtained at 13 t 1°C. 24 429. 308328135 Til I - o o O 3 2 I. TI. 2558 ere: 53:8er Eludni Volume (ml) Ac 160 I40 120 I00 80 60 40 20 180 PPM 25 concentration of the aldononitrile phosphate solution used for hydrogenolysis varied between 50 and 100 mM. Table 2. Purification and Epimeric Distribution of Aldononitrile Phosphates chromatography on Dowex 1-X8 parent aldose (formate)a ,, ratio of phosphates peak 1D peak’ZD epimersc glyceraldehyde-B-P threo erythro 1.3:1 erythro erythrose-4-P arabino ribo 1.4:1 ribo threose-4-P xylo lyxo 1.5:1 lyxo 3A 2.2 x 51 cm Dowex 1-X8 (200-400 mesh) column in the formate form was employed. Solutions of aldononitrile phosphates were adjusted to pH 6.5-7.0 prior to application to the column bed. Gradients: for 4-carbon aldononitrile phosphates, 3000 mL, 0.2-0.9 M sodium formate, pH 3.9; for 5-carbon aldononitrile phosphates, 3000 mL, 0.05-0.8 M sodium formate, pH 3.9. Temperature = 4°C; flow rate = 0.5 mL/min; 7 mL per fraction. Aldononitrile-P configurations were determined by reduction to gadose phosphates and incubation with alkaline phosphatase. The C NMR spectra of the resulting aldoses were compared with those of standard aldoses. cDetermined by computerized integration of 13C NMR spectra of epimeric mixtures and by quantitation of organic phosphate in purified preparations. Purified four- and five-carbon aldononitrile phosphate solutions were treated with 1 mL of glacial acetic acid per mmol of nitrile phOSphate and then with 2 M HCl or 2 M NaOH to pH 1.7 2 0.1. Reduc- tions were carried out as described for the three-carbon homologue. Aldononitrile phosphates were typically reduced for 6-8 h at 25°C. Incomplete reduction was noted in a few instances. In these cases, the spent catalyst was removed by filtration through Celite, and a second 26 reduction was performed to complete the conversion to the aldose phosphate. Hydrogenolysis products were assayed by 13C NMR (Figure 28 and 2C) to determine the extent of reduction to 1-amino-1-deoxyalditol phosphates and the amount of unreacted aldononitrile phosphates. Yields of the three, four-, and five-carbon aldose phosphates based on the analysis of 13C NMR spectral peak areas in reactions with [1-13CJ-a1dononitri1e phosphates are 85-95%. After hydrogenolysis, the catalyst was removed by filtration through Celite. The solution was treated with excess Dowex 50(H+), and after filtration was concentrated to 10 mL. Typically, the reac- tion mixture contains product aldose phosphate, I-amino-I-deoxyalditol phosphates, and a small amount of aldononitrile phosphate. This solu- tion was adjusted to pH 4.5 x 0.1 with dilute NaOH and applied to a 1.2 x 50 cm DEAE-Sephadex A-25(acetate) column at 4°C which had been equilibrated with 0.05 M sodium acetate at pH 4.5 t 0.1. The column was developed with a linear acetate gradient (1500 mL, 0.05-0.8 M sodium acetate, pH 4.5 i 0.1). Fractions (6 mL) were collected with a flow rate of 0.5 mL per min. The I-amino-I-deoxyalditol phOSphate eluted at the void volume, followed in order by aldose phosphate and the aldononitrile phosphate. Fractions containing aldose phosphate were pooled, treated with excess Dowex 50(H+) and concentrated in vacuo at 30°C to approximately 10 mL. Recovery from DEAE-Sephadex chromatography based on phosphate assay was greater than 95%. Aldose phosphate solutions were stored.at pH 4.0 and -15°C. Characterization of Aldose Phosphates. Purified 1-13C-enriched aldose phosphates (100 pmol) were incubated in 2 mL 27 of 50 mM Tris HCl buffer (pH 9.0) with alkaline phosphatase for 1 h at 36°C. The 13C NMR spectra of the resulting aldoses were compared with standard Spectra (Table 2). Ketose Phosphates. D-[2-13CJFructose-1,6-P2 and D-[1-13CJfruc- tose-1,6-P2 were prepared by the action of hexokinase, glucose 6-P isomerase, fructose-6-P kinase, and myokinase on D-[2-13Cngucose and D-[1-13C]glucose, respectively (Serianni, 1980). D-[2-13CJFructose-6-P was prepared from D-[2-13CJfructose 1,6-P2 by the sequential action of acid phosphatase and hexokinase (Midelfort et al., 1976). Fructose-6-P (0.5 mmol) was purified by anion exchange chromatography on a DEAE Ae25 Sephadex column (2 x 50 cm) that had been converted into the bicarbonate form. A 3 L linear gradient of 0-0.4 M NaHCO3 at pH 7.5 was used to elute the ketose phosphate. Fractions containing the ketose phOSphate were pooled, treated with Dowex 50(H+), and after filtration, were concentrated in .gggug at 30‘C. The pH was adjusted to 4 with 0.5 M NaOH, and the solution was stored frozen at -20°C. D-[1-13C]Xylulose-1,5-P2 was prepared from D-[1-13CJfruc- tose-1,6-P2 and glycolaldehyde-P by the action of fructose-1,6-P2 aldolase. The ratio of glycolaldehyde-P to D-[1-13C3fructose- 1,6-P2 was 20. Column chromatography was according to Byrne & Lardy (1954). D-Xylulose-1,5-P2 emerged from the column just after D-fructose-1,6-P2. Fractions containing D-xylulose-1,5-P2 were pooled, neutralized with 0.5 M NaOH, and a three fold molar excess of barium acetate was added. The bisphosphate was precipitated as its barium salt upon the addition of one volume of 95% ethanol and 28 storage at -20°C for at least 1 h. Centrifugatioh yielded a pellet that was suspended in H20 and treated with Dowex 50 (H+). After the aqueous phase had become clear, the solution was filtered and brought to pH 6 with 0.1 M NaOH. The filtrate was concentrated at 30°C 11129.29 and stored frozen at -20°C. D-[2-13C]Ribulose-1,5-P2 and D-[I-13CJribulose-I,5-P2 were prepared from D-[2-13C1ribose-5-P and D-[1-13CJribose-5-P, respectively, by the action of ribose-S-P isomerase and ribulose-S-P kinase (Horecker et al., 1958). 6-O-Methyl-D-[2-13C]fructose-1-P was prepared by the action of fructose-1,6-P2 aldolase and triosephosphate isomerase on 3-O-methyl- D-glyceraldehyde and D-[2-13CJfructose-I,6-P2. 3-0-Methyl-D-- glyceraldehyde was synthesized from methyl-(S-O-methyl)-D-riboside (Levene & Steller, 1933) which was further purified by chromatography on Dowex-1(OH') (Austin et al., 1963). Elution of the colwnn with deionized, distilled H20 resulted in three peaks (phenol-sulfuric acid assay; Hodge & Hofrieter, 1962), the second and third of which contained methyl-(S-O-methyl)-a—D-ribofuranoside and methyl-(5-O-- methyl)- a-D-ribofuranoside, respectively. These peaks were pooled, adjusted to 0.04 N in HCl, and heated under reflux for 2 h to produce 5-O-methyl-D-ribose. Treatment with Dowex-1(acetate) and concentration jn_ygggg_gave a deionized solution. The purified 5-0-methyl-D-ribose was treated with lead tetraacetate in the manner described for the preparation of O-glyceraldehyde-3-P from D-fructose-6-P. The product of the lead tetraacetate oxidation, Z-O-formoyl-3-O-methyl-D-glyceral- dehyde, was converted into 3-O-methyl-D-glyceraldehyde by incubation with 0.1 N H2504 at 35°C for 4 h. The acidic solution was 29 deionized by successive treatment with Dowex-1(acetate) and Dowex- 50(H+). D-[2-13CJFructose-l,6-P2 (620 umOI) and 3-O-methyl-D-- glyceraldehyde (5 mmol) were incubated with 500 units of aldolase and 1000 units of triosephosphate isomerase at pH 7.2 in a total volume of 120 mL. The reaction was judged to be complete when 25 pL of the solu- tion no longer caused a decrease in absorbance at 340 nm upon its addition to a solution (975 uL) containing aldolase, glycerolphosphate dehydrogenase, and 0.1 mM NADH. (It appears that aldolase has a high Michaelis constant for 6-0-methyl—D-fructose-1-P. Therefore, under the conditions of the assay, D-fructose-1,6-P2 and the triose phosphates will promote oxidation of the NADH, while 6-O-methyl-D-fructose-1-P will not.) Purification was accomplished by anion exchange chroma- tography in the manner described for the purification of D-fructose- 6-P. The presence of triosephosphate isomerase and aldolase in the synthetic reaction caused the 13C label to be diluted to 45 atom % .by equilibrating the isotopes at C-2 and C-4 of D-[2-13C1fructose- 1,6-P2. Ketose phosphates were analyzed by 13C NMR, enzymatic, and phosphate analyses. They were judged to be at least 95% pure. Results and Discussion The high yields and relative simplicity of the manipulations involved in the syntheses reported here make many of the biologically important sugars with terminal phosphate esters more readily available for use in biological systems. The syntheses can be applied using 13C- and/or 14C-enriched cyanide to afford the respective 30 13C- and/or 14C-enriched sugar phosphates. In the present study, the four diastereoisomeric pentose-S-phosphates have been prepared with 13C enrichment at C-2 by starting with D-glyceraldehyde-B-P and using two cycles of condensation and hydrogenation. The overall yield based on D-glyceraldehyde-B-P or K[13C]N was approximately 55%. Thus, serial application of the synthesis starting with glycolalde- hyde-P can presently be used to prepare aldose phosphates with enrich- ment at all but the penultimate and terminal carbons. The ketose phos- phates may be prepared by enzymatic modifications of the aldose phos- phates or their dephosphorylated derivatives. Assignment of Chemical Shifts (Table 3). Assignment of the chemi- cal shifts for C-1 of the a- and s-furanose phosphates was made by analogy to those assigned to the o- and e-methyl furanosides (Ritchie et al., 1975; Gorin & Mazurek, 1976). The C-2 chemical shifts are assigned on the basis of their 13C-13C coupling to C-1 in 1-13C-enriched compounds, or by direct observation of the 2-13C-en- riched compounds. Assignments for the carbons bearing the phosphate group and adjacent carbons are based on the predictable coupling pattern of phOSphorous to carbon (Lapper et al., 1973; Lapper & Smith, 1973). Linear aldose phosphates have C-1 resonances at approximately 91 ppm that are characteristic of linear, em-diol carbons (Grindley et al., 1977; Serianni et al., 1979b). Linear ketose phosphates also exhibit similar resonances, although highly de-shielded resonances characteristic of non-hydrated, keto carbons in the region of 210 ppm predominate. The measurements of chemical shifts and 13C-IH, 13C-31P, 13C-13C coupling constants can be used to determine the solution conformations 31 .mc_—a:ou amalgam soc; m:_m_cc um—nzou a me mcmmaaa oucmcomozu coo." H vm am $69: mLmZ mcowumcwELmeU mmmshu .m_ma—m:o mzz op Loren usmm an Au_== ~.owv :a co>_m ozu on cmpmznum use: mcopp:_omn .vogamuma Ho: ecu: cmcmucm go: «so gupgz mappzm paupsmgo .N: oocw mo swap: cmu__$ a ten N: coma mo zuupz ammzm a saw: uoH w mg an uocsmowe mew: mucwgm —mu_Em;um e.om m.m caucus; ea.me ee.om a.o~ m.- m.em m.m omeeecec-e e~.me ed.om ~.- ~.m~ m.HoH m.m omoeacec-e e a a-m-omoxmp-so .o m.m ounce e e~.me em.fim “.mc m.Hm o.mo~ m.m_ omeeecec-a e~.me ee.mc a.mc m.e~ m.em m.m omoeecec-e N a-m-omo_»x-so .oa m.m ounce e ee.ec em.~m N.H~ e.e~ e.~o~ m.m omoeecec-e em.me ee.mm m.HN m.- m.~m m.m onoeecec-e N Hm a-m-onom_c-o . m.m oeece e e~.ec e~.Hm H.m~ c.- m.ea m.m omoeacse-e eH.me e~.mm ~.e~ ~.~m «.moH m.m oneeecec.e aim-omo=_nmeM1o ec.~e ee.o~ H.e~ m.om m.e ounce»; .a-e-onoocee-do eo.me ee.- m.m~ m.om m.“ mecca»; .e-e-enoceezco-so eo.ee eo.e~ m.~¢ m.~ ooeece»; .e-m-oe»eoepacoo».m-o e~.me e~.om m.~ oooeceae .e-oe»eoee_ooa_m m-u 4-0 m-u ~-u H-u are eeeeeeoo uaaaav :oFBVmom concuo mopeeameee omee_< co mec_em _eo_Eoeo Una .m o_eec 32 of sugar phosphates, and are useful analytical parameters. Thus, for B-D-ribose-S-P and the D-ribose-S-P moieties in 5'-UMP and 5'-AMP, 3JPOCC = 8.4 Hz (pH 5.5), 8.5 Hz (pH 6.3), and 8.7 Hz (pH 6.3), respectively. The magnitude of these couplings indicates that the preferred position of the phosphate group is trans to C-4 and gauche to H-5' and H-5" (Alderfer & T50, 1977). Additionally, lac] H] couplings for the ggmrdiolic aldose phosphates range from 160 to 164 Hz. Formation of the furanose ring increases the value of 1JCl,Hl by about 10 Hz. This coupling can therefore be used to distinguish between ggmrdiol and furanose resonances when these resonances have similar chemical shifts. Solution Structures of Linear Sugar Phosphates. Glycolaldehyde-P and glyceraldehyde-3-P appear to exist predominantly as monomeric hydrates (linear ggmediols). Resonances at approximately 205 ppm which would represent aldehydo forms were not observed under conditions of signal to noise that should have allowed detection of under 5% of these forms. When the acid form of DL-[l-13C]glyceraldehyde-3-P was con- centrated to dryness, a mixture of dimers and oligomers with C-1 derived chemical shifts at 105.2, 103.8, 102.9, 97.3, 93.1, 92.8, 89.7, and 89.0 ppm were formed. These forms reverted over a period of time to the monomeric hydrate in dilute (0.1 M) aqueous solution at pH 1-2. Dilute aqueous solutions of DL-glyceraldehyde-B-P contain about 5-10% of these higher structures at pH 5.5. The tetrose-4-phosphates also exist principally as monomeric hydrates in dilute aqueous solution at pH 2-5, but chemical shifts at 97.9 and 98.4 ppm indicate the presence of up to 15% of dimers and/or oligomers in these solutions. 33 .uoH H on an cm:_cpao ago; mcuomnm .m:_—n:oo omfiiafim x9 nmmzco m_ c mmocmcommc mmmzu mo m:_uu_—Qm .scom mumccasuaumx z» E m mmue:_m_co sag mm xpmums_xocaam an um_n:ov use mmmcmzz .scow oumx mgu soc; mmumc_m_eo sag oHN »_wuma_xoeqae ea »m_naon asp .cowpcu_»_c=a Loewe czozm m_ ssepomam «z: u H mmozz Noim.fiuwmo_:nwcmu -mgio mmoacoca m o» mh:ou o_uaax~:m one .m wzszm 34 8 oo. o: om. omm ‘1 q q u — u — a — c o n.._.< + omo:_xo_3tocamozn_+ x x nag 3x m 6335383123351. e x x E < 4n 8.er 86-9-8. 35 The linear ketose phosphates exist primarily as the unhydrated, keto form in aqueous solution. Thus, the addition of ribose-S-P isomerase to D-[2-13CJribose-5-P causes the appearance of a down- field resonance (213.7 ppm) (Figure 3A, 3B) which is characteristic of the free keto form of D-[2-13CJribulose-5-P. The equilibrium favors the aldose phosphate (72%) at 36°C as previously observed by Axelrod & Jang (1954). Addition of ribulose-S-P kinase and MgZ+-ATP converts the downfield singlet into a doublet arising from 13C-31P coupling and shifts the equilibrium toward the product, D-[2-13C]ribulose-1,5-P2. The 13C NMR spectrum of purified D-[2-13C] ribulose-1,5-P2 (Figure 3C) at pH 7.6 shows doublets centered at 211.7 ppm (88%, 3JPOCC = 7.3 Hz) and 97.6 ppm (12%, 3JPOCC = 6.6 Hz) indicating that aqueous solutions of D-ribu- lose-1,5-P2 at pH 7.6 contain 88% keto and 12% hydrated forms. This result compares favorably with a determination made by infrared spectroscopy (Gray & Barker, 1970). D-Xylulose-1,5-P2 demonstrates similar behavior. Glycolaldehyde and Glyceraldehyde Phosphates. For the preparation of millimolar quantities of aldose phosphates, glycolaldehyde-P and D-glyceraldehyde-3-P were prepared from DL-glycerol-l-P and D-fructose- 6-P, respectively, by lead tetraacetate oxidation. Sodium metaperio- date oxidation was also examined, but traces of iodate interfere with hydrogenolysis of the aldononitrile phosphates, and careful chromatog- raphic purification was required. Aldononitrile phosphates prepared from lead tetraacetate oxidation products hydrogenolyze smoothly. Ion-Exchange Chromatography and Stability of Products. Epimeric tetrose and pentose phosphates are difficult to separate by ion-ex- 36 change chromatography (Bartlett, 1959). For this reason, the epimeric aldononitrile phosphates were separated on Dowex 1-X8(formate) at 4°C by using linear formate gradients at pH 3.9. Epimeric mixtures of these aldononitrile phosphates can also be separated on Dowex 1-X8 (chloride) using linear chloride gradients, but hydrogenolysis in the presence of high concentrations of chloride ion consistently yielded a larger amount of l-amino-I-deoxyalditol phosphates (~45%). It is important to maintain acidic conditions during the separa- tion and handling of aldononitrile phosphates since the reaction between the parent aldose phosphate and cyanide is reversible, and, at pH > 8, purified aldononitriles revert to epimeric mixtures. Aldose phosphates, particularly the triose and tetrose phOSphates, should be handled at low pH to avoid base-catalyzed isomerizations and B elimination. The acyclic triose and tetrose phosphates isomerize to give mixtures which include keto compounds when chromatographed on Dowex 1-X8(formate). Purification of the alkali-sensitive aldose phosphates and the pentose phosphates can be achieved by anion-exchange chromatography on DEAE-Sephadex A925 at 4°C by using linear gradients of acetic acid at pH 4.5 t 0.1. The tetrose-4-ph05phates consistently yielded skewed peaks with notable tailing, whereas triose and pentose phosphates yielded symmetric peaks. Isomerization to keto compounds on DEAE-Sephadex was not observed under theconditions used. The pentose phosphates, particularly xylose-5-P and lyxose-5-P, have also been seen to isomerize to keto compounds in aqueous solution at pH 6-8. It is usual to prepare aldose phosphates having four or fewer carbons as acetals to protect the base-sensitive aldehydic function (Ballou & Fischer, 1955; Ballou et al., 1955). Linear ketose phos- 37 phates are also base sensitive. It appears, however, that these sugar phosphates are stable during long term storage at pH (2. When stored at -15°C as 50 mM solutions, no detectable changes occur over a 2 month period as determined by 13C NMR analysis of the 13C-enriched compounds and inorganic phosphate analysis. Repeated freezing and thawing, however, is not recommended, as these manipulations have been observed to lead to dimer and oligomer formation. Summa y. The specific 13C enrichment of the compounds described in this chapter provides another means for studying their solution behavior. Some of the solution forms of the linear sugar phosphates have been described. Many of these descriptions corroborate earlier work using other techniques, and they demonstrate the utility of 13C NMR analyses of the enriched compounds. 0f greater imme- diate importance is the analysis of the cyclic sugar phosphates. The solution behavior of these compounds does not readily lend itself to analysis by conventional techniques. For instance, the 1H NMR spec- trum of D-arabinose-S-P cannot be interpreted by a first order analysis even at the highest-operating frequencies presently available (600 MHz) due to the complications of 1H-lH and 1H-31P couplings and degenerate chemical shifts. However, the 13C NMR spectra of the cyclic sugar phosphates are easily interpreted and easily obtained. This technique was therefore used to investigate the chemistry of the cyclic, furanose phosphates. The research is described in the next chapter. CHAPTER TWO THERMODYNAMIC AND KINETIC MEASUREMENTS OF THE TAUTOMERIZATION OF FURANOSE RING SYSTEMS Previous studies on the tautomerization of sugars and sugar phosphates (see Literature Review) have been mainly concerned with the rates of interconversion between the cyclic hemiacetal forms. It is the purpose of this chapter to describe some of the kinetics of tautomerization of furanose phosphates in terms of the unimolecular ring opening and ring closing rates. Furanose ring systems were chosen for study since these systems were expected to contain a larger amount of the acyclic carbonyl form than the pyranose systems. This expectation was confirmed, and the significant concentration of the acyclic carbonyl form made it possible to determine, for the first time, the thermodynamic parameters of the tautomerization reaction. Extensive use has been made of 13C NMR analyses of the 13C enriched compounds whose synthesis was described in Chapter 1. NMR Bandshape analyses and saturation transfer techniques were used to assess the kinetics of tautomerization. Theory The Tautomerization Reaction. The tautomerization reaction for furanose phosphates (lower half of Figure 1) may be viewed in a chemically reasonable fashion as 38 39 koo 2 kos a 1. 0 ... TI 3 (3) kOG kBO [HZOtho kho h where 0 represents the open, acyclic carbonyl form, h represents the hydrated carbonyl form, and a and 8 represent the cyclic furanoses. (The intermediacy of the acyclic, carbonyl form will be demonstrated later). The thermodynamic and kinetic relationships that apply to eqn 3 at equilibrium are shown in eqns 4 (Xw is the mole fraction of H20). xakao = x0km: 3 K60 = XO/Xa Xhkho ' onw oh ; Kho ‘ Xoxw Xh The NMR methods that will be used give the kinetic constants for ring opening directly. The kinetic constants for ring closure may then be determined by the relationships in eqns 4. It is often desirable to know the rate of interconversion between the cyclic furanose forms, a process which may be conveniently described by a simple first order reaction. a ,V_ “ B . (5) When the rate of interconversion between the open chain form and the hydrated form is much slower than the rates of interconversion between the open chain form and the cyclic forms (this will be demonstrated later), then the anomerization constants, kaB and kBa, may be 40 related to the unimolecular rate constants (eqns 3 and 4) by an exact solution to the rate equation of eqn 3 (Lowry & John, 1910; Los et al., 1956). The resulting equation is quite complex and contains squared and square root functions of the unimolecular rate constants. However, when the concentration of the acyclic carbonyl form is small compared to the concentration of the cyclic forms, a steady state kinetic analysis is suggested. Application of the steady state assumption to the concentration of the acyclic, carbonyl form results in considerable simplification of the exact solution. The result is presented as eqn 6. 08 (5) A comparison between the exact solution of eqn 3 and the steady state approximation (eqn 6) can be made using the data for D-fructose- 1,6-P2 given in Table 1. The value for the exact solution is k = 9.223 s-l, and for eqn 6, k = 9.243 s-l. The difference between these two treatments (ca. 0.2%) is negligible with respect to the experimental error in determining the rate constants. Therefore, eqn 6 will be used in the subsequent discussion of anomerization rates. Since eqn 5 implies that Xakae = kaBa ’ (7) the use of eqns 6 and 7 allow the determination of the anomerization constants, kaB and k8“. 41 NMR Linewidth Studies. It is well known that when chemical species undergo exchange, dramatic changes occur in the NMR spectrum depending on the exchange rate and the frequency separation of the resonances corresponding to the chemical Species (Gutowsky & Saika, 1953). These spectral changes result from the change in the lifetimes of the excited states of the nuclei that is caused by the chemical exchange. As the rate of exchange increases, the lifetimes decrease. Since the lifetime has become more accurately determined, the energy of the excited state has become less accurately determined according to the Heisenberg Uncertainty Principle. The nuclei correspondingly resonate over a broader frequency domain with the result that the linewidths of the resonances increase. For the reaction k1 S.;::::: I (8) k-1 which implies Xskl = X1k-1, (9) discrete resonances corresponding to S and I are observed when k1=k-1=0. As k1 and k-1 become positive, separate but broadened lines are observed in the original vicinity of the S and I resonances. As the rate of chemical exchange increases further, the resonances coalesce. Finally, in the limit of very fast exchange (when (k1 + k-1) is much greater than the frequency separation between the resonances), the species are interconverting much more rapidly than the sampling rate of the NMR frequencies. In this limit, one single resonance is observed at a position corresponding to the weighted average of the original positions of the S and I nuclei. 42 Detailed mathematical descriptions of these phenomena have been obtained (Gutowskyla Holm, 1956;McConnell,1958; Binsch, 1975). The general solution to the steady state NMR equations modified for chemical exchange is quite complex and requires computer modeling to derive the rate constants. However, for the cases in which XS<m mcu mew mama och .uoH w mm was N.~ :a pm mmumzamoca mmoumxmu Himg use mmumzamoza omen.8-Hu -Hu co eceooem 122 u a mm: m .mH 6;» Le m_n»_aee mamgmucmn x9 uw:_scmuwm acmz mono; mcwcmao m:_c ash .mmuwcamoga «mapmx new mmoupm as» :mean ma_;m:o_ym_mc u_um:wx can _ec:uu:spm .8 mzaumm 52 on OS wd a. N.N Em MN 9. 2.-.. 8.. x w 886.31% 0 . oa N.Q 00$ 06 . _ m n. NM N? :13 x m o m o m t o o m o m o m o m r 20.5 2 0:32.. 2: 9;. on... Quince... Quinn. 53 acyclic carbonyl form while the ketose phosphates contain a rather appreciable amount. In addition, the acyclic aldehydes are appreciably hydrated whereas the acyclic ketones are not. This behavior was also seen for the aldose and ketose phosphates that are constrained to be Table 4. Tautomeric Composition of Selected Furanose Ring Systemsa Percent in Solutionb Sugar 8 a hydrate free carbonyl D-ribose-S-P 63.9 35.6 0.5d 0.1c D-arabinose-S-P 40.4 57.3 2.2 0.2c D-xylose-S-P 42.4 52.6 4.7 0.3d D-lyxose-S-P 24.8 70.5 4.3 0.4d D-fructose-6-P 81.8 16.1 e 2.2 D-fructose-1,6-P2 86.0 13.1 e 0.9 D-erythrose 63.8 21.1 14.4 0.7 aDetermined by 13C NMR analysis of the 13C enriched compounds (0.4 M in 15% [2 H120) at 15. 08 MHz. T= 6 i 1°C, pH 4. 5. The error in the measurements is (8% of the value given, unless otherwise noted. cThese values may be in error by as much as 50% due to the very low amounts in solution. dThe error in these values is (15%. eResonances for these forms were not observed. 54 acyclic. That is, D-erythrose-4-P was greater than 90% hydrated in aqueous solution, whereas D-ribulose-1,5-Pz was only hydrated to the extent of 12% (Chapter 1), even though the electron withdrawing properties of the phosphate group should render the carbonyl carbon of D-ribulose-1,5-P2 more electrophilic (Bell, 1966). These data remind us that aldehydes are generally more susceptible to nucleophilic attack than are ketones, and therefore, the aldoses contain higher proportions of the cyclic and hydrated carbonyl forms than do the ketoses. D-Erythrose demonstrates anomalous behavior. For instance, the acyclic modifications account for almost 15% of the total sugar in D-erythrose solutions, while accounting for (1% of the total sugar in solutions of D-ribose-5-P. Increased ring strain in the tetrofuranose ring as compared to the pentofuranose ring seems to be an unlikely explanation. The observed differences in the relative amounts of the cyclic and acyclic carbonyl forms may reside in the expectation that the secondary alcohol of D-ribose-S-P (OH-4) should be a stronger base than the primary alcohol of D-erythrose (OH-4), though there is no direct evidence to support this explanation. D-Threose demonstrates behavior similar to that of D-erythrose (data not shown). The relative concentrations of the acyclic carbonyl forms were determined at low temperature and pH 4.5, since at higher temperatures and higher pH, interconversions between the tautomers would become more rapid and the subsequent linebroadening of the NMR signals would result in an unacceptable loss in signal to noise. However for two of the compounds, D-erythrose and Defructose-6-P, the linebroadening effects were small enough to allow the determination of the temperature dependence of the tautomeric equilibrium constants. From the observed 55 temperature dependence, the thermodynamic parameters for the tautomeri- zation reaction could be calculated, as demonstrated in the next section. Thermodynamics of Tautomerization. The tautomeric equilibria of D-erythrose and D-fructose-6-P are temperature dependent. Examples of representative spectra from which this conclusion was derived are given (Figure 5A) to illustrate the excellent resolution and signal to noise of the 13C NMR spectra of the 13C enriched compounds. The relative amount of the acyclic carbonyl form increases with increasing temperature for these compounds (Figure 58 and 5C), changing approxi- mately 3 fold over a 40°C temperature range. From these data, the thermodynamic parameters for the tautomerization reaction may be deter- mined (Table 5). Although unknown contributions to these parameters from solvation effects make detailed analysis difficult, the possession of these quantities allow the following conclusions. In all cases, the heat of reaction makes a larger contribution to the equilibriwn posi- tion at 253C than does the entropy of reaction. Formation of the more stable cyclic form for a given sugar is attended by slightly larger enthalpic and entropic values than for the formation of the less stable cyclic form. Whether this property is due to differences in internal energies of the molecules or to differing extents of hydration is unclear, although the much less negative entropy values obtained for the cyclization of D-threose in pure [ZHJZO (A.S. Serianni, per- sonal communication) suggest that hydration of the cyclic forms may contribute a great deal to the observed entropy of reaction. The thermodynamic values obtained for the aldehyde hydration reaction of D-erythrose are comparable to those obtained for other aldehyde-hydrate 56 FIGURE 5. Temperature dependence of the tautomeric equilibria in solutions of D-erythrgse and D-fructose-6-P. (A) The 15.08 MHz 13C NMR spgctra of D-[2-1CJfructose-6-P (at 42 1 1°C) and D- [1 CJerythrose (at 36 1 1°C) are presented as representative spectra. The a- and a-furanose forms are the major species along with resonances corresponding to the free carbonyl forms (0), hydrated carbonyl form (h), an linear keto form (k). Unmarked resonances in the spectrum of D- [1- 3C]- erythrose are due to dimeric and oligo- meric forms. B) The temperature3 dependence of the tautomeric equilibria in golutions of D- [2 3C]fructose-6-P at pH 4.5 (0.4 M sugar in 1135% [ H] 20). The equilibria were determined by integra- tion of1CNMR spectra at various temperaturesH and are the average of three determinations at each temperature. = 4. 8+ 0.5 kcal/mol; = 3.8 1 0.4 kcal/mol. (C) The AHgemperatprei depen- dence of the “tautomeric equilibria in iolutions of D-[13CJery- throse at pH 4.5 (0.4 M sugar in 15% [ HJZO). The equilibria were determined by integration of C NMR spectra at various tempera- tures and are the average of three determinations at each temperature. AH§,0 5.7 1 0.6 kcal/mol; M390 = 4.6 1 0.5 kcal/mol; AH,»0 5.3 1 0.5 kcal/mol. fructose 6-P ,Lx. 57 erythrose 1 18C) 140 ppm Figure 5A ln(Keq) 4.0 ' 3.0 2.0 58 F6P (B/Open) ( a/open) / 3.2 3.4 3.6 (l/T)x|O3 Figure SB In(Keq) 4.0 3.0 2.0 59 Erythrose (hydrate/open) l l 3.2 3.4 (I/T)x|O3 Figure BC 3.6 60 Table 5. Thermodynamic Parameters for the Tautomeric Equilibria of D-Erythrose and D-Fructose-6-P at 25°C3 aoob 1H0 65 Sugar Reaction (kcal/mol) (kcal/mol) (cal/K-mol) D-erythrose B-open 2.2 5.7 11.6 a'Open 106 406 908 hydrate-open 1.4 5.3 13.2 D-fructose-6-P B-open 1.8 4.8 10.2 a-Open 009 308 ' 905 aDetermined by 13C NMR analysis of the 13C enriched compounds in 15% [2 H120 at 15. 08 MHz. Temperature range: 45°C; pH 4. 5; sugar] = 0.4 M. Values are based on the average peak areas for three spectra at each of 4-5 temperatures and are accurate to within (10%. Standard state= 1 M for all components, except H20, which is arbitrarily assigned an activity of 1. equilibria (Bell, 1966). The determination of the tautomeric equilibria and the effect of temperature on these equilibria are important for the kinetic studies of tautomerization that will be discussed presently. NMR Linebroadening Studies on the Kinetics of Tautomerization. The use of bandshape analysis for the determination of chemical exchange rates has been discussed in the Theory section. The line- widths of the anomeric resonances of the aldose and ketose phosphates are strong functions of pH as shown for D-[1-13C]ribose-5-P (Figures 6A and 6B). Eqns 11 were used to convert the observed line- widths into ring opening rates. Although the small amount of line- broadening at pH 2-4 prevented an accurate determination of the ring opening rates in this pH range, it is clear that increasing the pH increased the ring opening rates. The ring opening rates were not a 61 FIGURE 6. The pH dependence of the ring opening rates for D-ribose- 5-P- (A) The 15.08 MHz NMR spectra of D-[1-13C1ribose-5-P (0.3 M sugar in 15% [2H]20) were obtained at 28 1 1°C. The sharp resonance on the downfield side of the a resonance is due to an unidentified component. A line broadening exponential factor of 0.4 Hz was used in transforming the spectra. (B) The ring Opening rate constfnts were calculated from the 15.08 MHz 3C NMR Spectra of D-[l- 3CJribose-S-P at 24 1 1°C, and are the average of three determinations. The error in the rate constants at pH > 4.5 is < 8%. 62 Figure 6A 63 20 30 IO Ms") Figure 6B 64 function of the concentration of D-[1-13CJribose-5-P in the range of 0.05-0.3 M at pH 5.5, indicating that intermolecular catalysis, if present at all, is dominated by intramolecular effects. It appears that the ring opening is catalyzed by the phosphate group, as there is a linear relationship between the observed rate constants and the ionization state of the phosphate group, both of which have apparent pK's of approximately 6.1. This behavior has also been seen with D-glucose-6-P (Bailey et al., 1970). All of the sugar ph05phates that were analyzed showed similar behavior, although the catalytic effi- ciency of the phosphate group appears to depend on the configuration of the furanose ring (Figure 4; Table 6). For the three ketose phos- phates, k80 > koo' It is interesting to note that the rate constants for 6-0-methyl-D-fructose-1-P are almost equivalent to those for D-fructose-1,6-P2. The differences may be more apparent than real. The use of D-[UL-13C]fructose-1,6-P2 in the studies of Midelfort et al. (1976) from which the rate constants for D-fructose- 1,6-P2 were obtained, resulted in considerable complication in the spectrum due to 13C-13C coupling. Overlap of the multiplets with the central resonance would cause the lines to appear broader and the rate constants to be over estimated. The large rates of ring opening for the dianionic forms of D-ribose-S-P and D-arabinose-S-P allowed the determination of the activation energies for ring opening (Figures 7A and 7B) of these two compounds. It can be seen that ring opening for D-ribose-S-P has a lower activation energy than that for D-arabinose-S-P. This difference is consistent with the greater catalytic efficiency of the phosphate group in the former compound. 65 |n(k) 4.0 *- .. 13 3'0 "Ea=|0.31 l.O kcol/mol ‘ 2.0 L‘ .1 l.() -' 1 3.2 31.4 316 (I/T) x 103 FIGURE 7. Arrhenius plots of the temperature dependence of the ring opening rates for (A) D-ribose-S-P and (B) D-arabinose—S-P at pH 7.5. The rgng opening rates were calculated frpm the linewidths of the 15.08 MHz 8 NMR spectra of solutions of D-[l— C]ribose-5-P and ' D--[1-l CJarabinose-5-P (0.2 M sugar, 15% [2H]20), and are the average of three determinations at each temperature. The rate constants are accurate to within 1 8%. ln(k) 4.0 3.0 2.0 LC 66 B/ Ea=|6.4 t |.3 kcol/mol l l Ea= 14.9: 1.2 kcol/mol 3.2 3.4 (I/T) x 103 Figure 78 vi 3.6 67 Table 6. Ring Opening Rates for Furanose Phosphates at pH 7.2 and 28°Ca Sugar kBo(S'1) kaO(S-1) ribose-5-P 22 42 arabinose-S-P 6.6 11 xylose-5-P 2.2 7.3 lyxose-S-P 3.7 3.7 fructose-6-P 7.0 6.0 6-O-methyl-fructose-1-Pb 35 8.2 fructose-1,6-P2c 45 9.6 aDetermined by Eineshape analysis of the 13C enriched compounds (0.2 M in 15% [ H320 at 15.08 MHz (see text for details). The values given are from the average of three spectra that were obtained with a digital resolution of 0.12-0.24 Hz/computer point and a linebroadening factgr of 0.4 Hz, and are accurate to within 10%. bCorrected for 1 C- 1P coupling as described in Midelfort et al. (1976). cFrom Midelfort et al. (1976). The identity of the rate determining step in the tautomerization reaction is uncertain. It was of interest, therefore, to determine the solvent isotope effect on the ring opening rates. This experiment is made difficult because, as has been shown, the observed rates are strong functions of the extent of ionization of the phosphate group. Therefore, a determination of the solvent isotope effect is only valid if comparison is made between two solutions, one in H20 and one in [2H]20, that contain the sugar phosphate in the same state of ionization. As the ionization reaction of weak acids is known to be accompanied by a solvent isotope effect (Bunton & Shiner, 1961), a 68 method for obtaining sugar phosphate solutions of the same degree of ionization in H20 and [ZHJZO is needed. The solvent isotope effect on the ionization of weak acids increases linearly with increas- ing pK (Bunton & Shiner, 1961). By interpolation of the available data, the pK of the phosphate group is expected to increase by 0.3 pH units in [2H]20. Therefore, a solution of sugar phosphate with pKz = 6.1 will be 96% in the dianionic form at pH 7.5 in H20, while in [2H]20, pKz = 6.4, and the sugar phosphate will be 96% in the dianionic form at p(2H) = 7.8 (pH = 7.4). Under these condi- obs tions, the activation energies for the ring opening of both anomers of D-arabinose-S-P were found to be 15.9 1 1.5 kcal/mol in [2H120, which is similar to the activation energies found for these processes in H20. The absolute rates of reaction, however, were 2.1 fold lower in [2H320 than in H20. This difference cannot be due to an error in the assumption used above, that pKz = 6.4 in [2H120. That is, for the value 2.1 to be due solely to changes in the extent of ioniza- tion of the phosphate group would require that in [2H]20, pKz = 8, which is unreasonable. Therefore, the effect of substituting 2H for 1H in the solvent H20 is one of lowering the pre-exponential frequency factor in the Arrhenius equation. As has been mentioned, lack of appreciable line broadening in solutions of monoanionic sugar phosphates made an accurate determina- tion of the ring opening rates impossible. However, since it was con- sidered desirable to accurately determine the catalytic efficiencies of the phosphate monoanions and dianions, it was necessary to measure the ring opening rates for the monoanionic sugar phosphates. The technique of saturation transfer spectroscopy was employed for this purpose. 69 Saturation Transfer Spectroscopy. It is clear from the data in Table 4 and Figure 5A that under normal conditions of signal to noise the resonance due to the acyclic carbonyl form in solutions of the furanose phosphates will be lost in the noise. This is a severe problem for D-[1-13C]ribose-5-P and D-[1-13C]arabinose-5-P even at 100.58 MHz. Nevertheless, by accumulating a large number of tran- sients, the signal can be observed. Then, upon application of a selective frequency that fully saturates the carbonyl resonance, the resonances due to the cyclic and hydrated tautomers decrease in inten- sity with time of irradiation according to eqns 13 or 15. Representa- tive spectra that were obtained with this technique may be found in Figure 8A. Figure 8B demonstrates the excellent fitting of the data by eqn 13. By analyzing the data for the various sugar phosphates by a least squares analysis of the linearized form of eqn 13, or by direct curve fitting, the rate constants for the ring opening rates may be determined (Table 7). The utility of the saturation transfer method for determining the rate constants for ring opening of the monoanionic sugar phosphates may be demonstrated by noting that all of the sugar phosphates tested had ring opening rates of (0.9 s"1 at 40°C and pH 4.5, which would correspond to a linebroadening of (0.3 Hz. This is too small a change in the linewidth to be accurately determined by bandshape analysis, but is well within the range of the saturation transfer method. It had been hoped that the saturation transfer technique would allow the determination of the rates of hydration of the acyclic carbonyl tau- tomers. However, at 40°C and pH 4.5, the rates of hydration were too slow ((0.04 5‘1) to be accurately determined even by this method. 70 FIGURE 8. Saturation igansfer spectroscopic analysis of the ring opening rates of D-[I- C]arabinosea5-P at pH 4.5 and 40°C. (A) The 100.58 MHz 13C NMR spectra were obtained at various times after the saturating frequency was applied to the aldehyde resonance. The larger downfield resonance in each panel corresponds to the a-furanose form, and the upfield resonance corresponds to the s-furanose form. A linebroadening factor of 5 Hz was used in transforming the spectra. (B). The effect on the a- and 8-resonance intensities of increasing the length of time of saturation of the aldehyde resonance. The curves were calculated0 from eqn 13 with (for the a-anomer): p=0.46 5' =0 64 =1. 0; =0 42; (and for the B-anomer): p= .485 ,k 1= 00. 49 s-T; Io=0. 66; 1,=0.33. 7I 0005s Ofils 160$ 2565 5775 I9405 1 1 I I __ 1401.211..qu #011011. Figure 8A 36:35 xoma 0.25 '- Time (s) Figure 8B 73 Table 7. RingaOpening Rates for Furanose Phosphates at pH 4.5 and 40°C Sugar kgo(5'1) kao(5'1) ribose-S-P 0.44 0.86 arabinose-S-P 0.49 0.64 fructose-G-P 0.22 0.20 aDetermined by saturation transfer spectroscopy at 100.58 MHz. [Sugar] = 0.15 M in 15% [2H2]0. Values are accurate to within 8%. Discussion A number of parameters pertaining to the solution behavior of sugar phOSphates have been determined. The use of specifically 13C-enriched compounds greatly facilitated the analysis of their solution reactions with presently available spectroscopic techniques. Tautomeric Equilibria. In the pentose phosphate series, the proportions of the linear tautomers are higher for those compounds with 0H-3 and C-5 in a cis arrangement (xylo- and lyxo-) than for those compounds with the corresponding trans arrangement (ribo- and arabino-) (Table 4; see also, Figure 4). These conclusions appear to be well correlated with a number of other solution properties. For instance, xylose-S-P and lyxose-S-P have been observed to isomerize quite readily to thrggfpentulose-S-P at pH>6. Up to 20% of the total sugar in these solutions has been found as the keto modification (data not shown). In 74 contrast, solutions of ribose-S-P and arabinose-S-P are much more stable to isomerization under the same conditions. This is reasonable behavior since the isomerization is believed to occur by enolization of the free aldehyde form. Isomerization is often accompanied by dimer and oligomer formation in lyxose-S-P and xylose-S-P solutions. Like- wise, the large proportion of the acyclic modifications in erythrose solutions engenders isomerization and oligomerization (Figure 5). All of these compounds are much more stable to isomerization at low pH, and their storage as dilute (to prevent oligomerization), acidic solutions is therefore recommended. An analogous isomerization of fructose-6—P to glucose-G-P and mannose-G-P might be expected due to the very large proportion of the acyclic ketone form in fructose-G-P solutions (Table 4). However, these isomerizations have not been observed even when fructose-G-P was kept at 0.3 M at pH 7.5 and 40°C for up to 12 hr. It appears then that enolization of the ketone function occurs much less readily then enolization of the aldehyde function. The study of the tautomeric proportions in pentose phosphate solu- tions is also important with regard to the solution behavior of the more numerous hexoses and their derivatives. Solutions of hexopy- ranoses and pentopyranoses are also in equilibrium with the correspond- ing furanose derivatives (Figure 1). Generally, the pyranose is the more stable structure, but the amount of the acyclic form will also depend on the concentration of the furanose modification. Now, the finding that the 5-0-methyl and S-deoxy derivatives of the pentoses have tautomeric proportions nearly identical with those of the corre- sponding pentose phosphates suggests that changing the substituents 75 at 0-5 of the furanose ring has little effect on the tautomeric equilibria. Therefore, the knowledge of the solution concentration of the furanose modification should allow the estimation of the amount of the acyclic carbonyl modification. For instance, altrose contains approximately 20% of the a furanose and 13% of the 3 furanose forms at 40°C (Angyal & Pickles, 1972). These forms have the arabino- configur- ation, with C-2 and C-3 EEEEE- From the data for D-arabinose-S-P in Table 4, and assuming that the temperature dependence of the tautomeric equilibria for D-arabinose-S-P are similar to those for D-erythrose and D-fructose-G-P (Table 5), one may estimate that solutions of altrose at 40°C should contain approximately 0.2% of the acyclic aldehyde form. Since it is much easier to measure the amount of the furanose form in solutions of the hexoses than it is to measure the concentration of the acyclic carbonyl form, the difficult problem of determining the amount of the acyclic carbonyl form in pyranose solutions is thereby simpli- fied. As many carbohydrate reactions are presumed to occur through the intermediacy of the acyclic carbonyl form, this knowledge should great- ly facilitate kinetic and physical analyses of carbohydrate deriva- tives. Further studies with the 5-0-methyl- and 5—deoxy-pentose derivatives should allow more accurate determinations of the temperature dependence of the tautomeric equilibria as the lack of appreciable linebroadening in their spectra results in more acceptable signal to noise levels. Kinetics of Tautomerization. The ring opening rates for the pentose and ketose phosphates appear to be linear functions (Figure 6B) of the extent of ionization of the phosphate ester moiety (pKZ m 6.1) between pH 4 and 7. At lower pH, acid catalyzed ring opening 76 occurs. At pH > 7.5, further increases in the ring opening rates suggest that processes other than intramolecular catalysis by the phos- phate group are occurring, such as specific base catalysis. However, intramolecular participation of the phOSphate group appears to be the predominant factor in controlling the ring opening rates throughout the intermediate pH range. The relative ring opening rates of the a- and B-furanose forms are of interest in regard to possible mechanisms of ring Opening which include catalysis of removal of a proton from OH-l and/or catalysis of protonation of the ring oxygen. Regardless of whether the aldose phos- phates are in the mono-anionic form (Table 7) or in the di-anionic form (Table 6; Figure 4), the a anomer opens at a faster rate than the B anomer. Since 0H-1 and the phosphate group are on opposite sides of the furanose ring in the a anomer, the direct participation of the phosphate group in the deprotonation of OH-l is precluded. Therefore, it appears that the phosphate group may catalyze the ring opening reac- tion by stabilizing the protonated ring oxygen. This suggestion is in accord with the observations that ring opening of the B anomer of the ketose phosphates is more rapid than that of the a anomer. The rela- tively lower rates observed for the asketofuranose phosphates and the B-aldofuranose phosphates may be rationalized by noting that in these cases, respectively, C-l or 0H-1 is in a cis-1,3 relationship with the C-5 phosphohydroxymethyl group. This interaction should tend to pre- vent the approach of the oxygen anion of the phosphate group to the ring oxygen, and a corresponding decrease in the ring opening rate constant is expected and observed. The finding that 6-0-methyl-D-fruc- tose-l-P tautomerizes nearly as rapidly as D-fructose-1,6-P2 77 (Table 6) suggests that the C-1 phosphate ester is completely competent in catalyzing the reaction. Accordingly, the lowest rate for the ketose phosphates is observed for D-fructose-6-P, which must overcome the unfavorable interaction of the C-5 phosphohydroxymethyl group with either C-1 or the anomeric hydroxyl group prior to catalysis by the phOSphate group. The relative stimulation of the ring opening rates and the anomer- ization rates by the phosphate mono-anion and di-anion depends on the ring structure of the furanose (Table 8). This dependence may also be seen by noting that the ring opening rates for a given anomer in the monoanionic forms differ by less than a factor of 4 (Table 7), whereas the dianionic forms may demonstrate larger differences (>10 fold) in ring opening rates (Table 6). The relative catalytic efficiences of the phosphate group therefore depend in a complicated way on the Table 8. Comparison of Tautomerization Rates for Selected Furanose PhOSphates at 40°Ca Rate Constants 13114.5k DHU750‘BE::$(_74_-_55)51301(T3)y Sugar k kBO as k co co ribose-S-P 0.86 0.44 0.41 100 40 44 120 100 arabinose-S-P 0.64 0.49 0.22 33 22 11 52 45 fructose-6-P 0.20 0.22 0.16 18 21 15 90 95 fructose-1,6-P 32 110 30 aRate constants were determined by linebroadening at 15. 08 MHz or saturation transfer spectroscopy at 100. 58 MHz, and are accurate to within 8%. Solutions were 0.15 M sugar phosphate in 15% [2 H]20. bDetermined by extrapolation from Figures 7A and 7B. cData from Midelfort et al., 1976. kaB was calculated from eqn 6 using the temperature dependence of the equilibrium constants for fructose 6-P (Table 5). 78 stereochemistry of the substituents at C-2 and C-3. This dependency may have its origin in the type of steric effects that were mentioned previously. A fuller set of data will be required before a more detailed statement is possible. It may also be noted (Table 8) that knowledge of only the anomeric rate constants, kaB and kBa’ may be misleading. For instance, the ring opening rate at pH 7.5 for a-D-ribose-S-P is 3.1 times greater than that for a—D-fructose-1,6-P2, although the rate of anomeriza- tion, kaB, is only 1.5 times greater than that for D-fructose 1,6-P2. Furthermore, for fructose-1,6-P2, kBa/kae = 0.15 whereas kso/kao = 3.4. Since mechanistic deductions must be made from the ring opening rates, the usefulness of the NMR techniques outlined in this chapter may be appreciated. The ring opening rates were not correlated with the thermodynamic stability of a given anomer (Tables 6, 7 and 8). For instance (Table 8), at pH 7.5 the more stable a anomer of D-arabinose-S-P opens at 1.5 times the rate of the B anomer; the less stable a anomer of D-ribose- 5-P opens at 2.5 times the rate of the B anomer; and the ring opening rates for the two anomers of 0-fructose-6-P are approximately equiva- lent even though there is a 5 fold difference in their thermodynamic stability. These observations again confirm the importance of knowing the unimolecular ring opening rates, since the use of the summed, anomerization rate constants might well provide opposite conclusions. Biological Implications of Tautomerization. A number of enzymes have been shown to be specific for a given tautomer of their substrate. Fructose-1,6-P2 phosphatase appears to be specific for the a anomer of 0-fructose-1,6-P2 (Benkovic et al., 1974) whereas phosphofructokinase 79 is specific for the 3 anomer of D-fructose-G-P (Fishbein et al., 1974). Therefore, depending on the metabolic pathway involved, one enzyme could produce as its product a particular tautomer although the next enzyme in the metabolic sequence might require a different tautomer. In such a situation, non-enzymatic tautomerization might well be rate limiting for some metabolic sequences (Schray & Benkovic, 1978). For instance, D-fructose-1,6-P2, D-fructose-6-P, and D-ribose-S-P are all involved in the photosynthetic carbon reduction cycle and participate in reactions catalyzed by fructose-1,6-P2 aldolase, fructose-1,6-P2 phosphatase, transketolase, and ribose-S-P isomerase. It is interest- ing to compare the rates of tautomerization of these compounds with the metabolic rate of the photosynthesis cycle. Steady state rates of photosynthesis have been observed in whole spinach leaves at the level of 20-60 nmol 002 fixed/s/mg chlorophyll at room temperature (Jensen & Bassham, 1966). D-Ribose-S-P and D-fructose-1,6-P2 are present in chloroplasts to the extent of approximately 3.2 and 4 nmol/mg chloro- phyll, respectively, under steady state Conditions (Lorimer et al., 1978). From the ring opening rates for D-ribose-S-P and D-fructose- 1,6-P2 given in Table 6, and the anomeric proportions in Table 4, one may use eqn 6 to calculate anomeric rates for these two compounds at pH 7.2 and 28°C [ch3 (ribose-S-P) ~20 s-l; kae (fructose- 1,6-P2) ., 9.3 s-1]. With equilibrium proportions of 36% a.- D-ribose-S-P and 13% a-D-fructose-1,6-Pz, these rate constants would correspond to i_,vivo rates of 23 and 4.8 nmol p anomer produced/s/mg chlorophyll, reSpectively. These rates are of the same order of magni- tude as the observed metabolic rate. Therefore, under appropriate conditions, the anomerization reaction may affect the partitioning of 80 different metabolic sequences. This conclusion is in accord with the analyses of possible effects of tautomerization on the glycolytic and gluconeogenic pathways (Schray & Benkovic, 1978). These workers also suggested that the presence of anomerase enzymes could change this partitioning and, if the anomerases exhibited allosteric responses to metabolites, then their modification of the rates of tautomerization could well represent an important control point in the regulation of metabolic activity. More detailed statements concerning the metabolic consequences of tautomerization will require a sure knowledge of the tautomeric rates and the specificities of the enzymes involved. Summa y. The use of NMR linebroadening and saturation transfer techniques for the study of the tautomeric reaction of a variety of sugar phosphates has been presented. The use of specifically 13C-enriched sugar phosphates greatly simplified the spectrosc0pic analysis. Thermodynamic parameters for two representative furanose ring systems were obtained along with the unimolecular ring opening rates of a number of furanose phosphates. Analysis suggests that intramolecular catalysis by the phosphate group increases the rates of ring opening, but that the rates are not vastly different than the rate of flux of metabolic sequences in which these sugar phosphates participate. CHAPTER THREE ALKALINE DEGRADATION 0F RIBULOSE-1,5-BISPHOSPHATE AND ITS EFFECT ON RIBULOSE BISPHOSPHATE CARBOXYLASE/OXYGENASE Sugar phosphates which have the phosphate moiety a or a to the carbonyl group are well-known to be unstable to alkaline treatment (see Literature Review). In this chapter, two of the products of base treatment of ribulose-P2 are identified and shown to be inhibitors of the enzyme, ribulose-P2 carboxylase/oxygenase. Materials and Methods The preparation of sugar phosphates was describédxin\Chapter 1. Ribulose-Pz carboxylase/oxygenase was prepared by standard procedures (Ryan & Tolbert, 1975), and its activity was determined with a radio- metric assay (Lorimer et al., 1976). The enzyme was incubated at 2 mg/mL for at least 1 h in the assay buffer (0.1 M N,N-bis(2-hydroxy- ethyl)glycine ("Bicine") at pH 8.0, 10 mM MgCl2, 0.2 mM Nag-EDTA) at 30°C. The enzyme solution was then adjusted to be 10 mM in NaHC03 and kept for 30 min at 30°C. Aliquots of this stock solution were added to the assay buffer which contained 20 mM NaH [14c103 (0.16 Ci/mol) so that the final protein concentration was 80 ug/mL (0.14 uM). The reaction was initiated by the addition of ribulose-P2 to a final concentration of 0.5 mM. After 20 to 60 s, the reaction was 81 82 terminated by the addition of HCl, and the acid stable radioactivity was determined by liquid scintillation counting. The specific activity of different enzyme preparations was from 1.0 to 1.4 nmol 002 fixed per min per mg proteins under these conditions. Protein determinations were performed according to a modified Lowry procedure (Bensadoun & Heinstein, 1976). Solutions of the pentulose bisphosphates were analyzed by gas chromatography as their acetylated pentitol derivatives. A ribulose- P2 or xylulose-Pz solution was reduced with NaBH4 at pH 7 and treated with alkaline phosphatase. The solution was deproteinized by centrifugation after the addition of hot ethanol. Boric acid was removed as methyl borate by repeated concentration from anhydrous HCl/methanol (5% weight/volume). Addition of water afforded a solution ‘which was treated with Dowex 50(H+). The compounds were then concentrated to dryness in vague, peracetylated with acetic anhydride, and subjected to gas-liquid chromatography (0V-225 column, 205°C). A control containing only the phosphatase enzyme was treated in an identical fashion. Results and Discussion Alkaline Degradation of Ribulose-Pp. Ribulose-Pz and xylulose-Pz lost inorganic phosphate at a high rate when incubated at 40°C and elevated pH (Figure 9). It was determined that, besides inorganic phosphate, the products of the phosphate elimination were not ribose-S-P or ribulose-S-P, since incubation of the base 83 4.0.. pH ILO 3.0- pH 8.0 O'RIHTT' ’2‘ E 20- cf pH 8.0: Mg” 1.0-J 0 I 1 T r 1 o _ 5.0 I0.0. Time (hrs) FIGURE 9. Time course of phosphate elimination from ribulose-Pg at 40°C. The concentration of ribulose-P2 was 4.6 mM. Experiments at pH 11 were maintained at that pH by the addition of NaOH through the reaction time. Experiments at pH 8 were buffered in 30 mM tris- (hydroxymethyl)aminomethane-HCT. The divalent metals were added as their chloride salts to a final concentration of 20 mM. 84 treated ribulose-P2 solutions with phosphoriboisomerase and phosphoribulokinase in the presence of ATP did not yield ribulose-P2 which could be detected by an enzymatic assay with ribulose-P2 carboxylase. when no further ribulose-P2 remained, the amount of inorganic phOSphate was 50% of the total phosphate, indicating that only one phosphate group from ribulose-P2 had been lost as inorganic phosphate. Since Mgz+ and Mn2+ are present in chloroplasts and are cofactors for ribulose-P2 carboxylase, the effects of these two cations on the stability of ribulose-P2 in basic solutions were tested (Figure 9). Whereas 20 mM Mgz+ did not enhance the rate of ph05phate elimination at pH 8.0, 20 mM Mn2+ (a large amount of red-brown precipitate was formed, presumably manganese oxides) enhanced the rate considerably. It is not known whether the presence of these two cations resulted in the same product formation. Also, when incubated under the exact conditions of the carboxylase assay (see Materials and Methods) but with the omission of ribulose-P2 carboxylase, ribulose-P2 lost inorganic phosphate at a linear rate (over 40 h) of 6.25 uM/h, which corresponded to a 1.3% loss of ribulose-P2 in 1 h. The significance of these data with regard to the kinetics of ribulose-P2 carboxylase will be discussed later. To investigate the mechanism of ribulose-P2 and xylulose-PZ breakdown, ribulose-P2 was incubated at room temperature and pH 12 and the absorption spectra recorded as a function of time. Initially, an absorption at 430 nm rose with time and then was partially bleached. Furthermore, addition of concentrated HCl to lower the pH to 2 caused 85 complete bleaching of this absorption. This type of absorption change is characteristic of a—dicarbonyl compounds (e.g. 2,3-butanedione and phenylglyoxal) which have characteristic absorption spectra including bands in the region of 400 to 460 nm (Scott, 1964). The magnitude of this band for 2,3-butanedione is pH dependent as base enhances the extinction coefficient. Presumably, the bleaching of the 430 nm absorption at long times is due to a breakdown of the initial products formed from ribulose-P2 upon loss of inorganic phosphate. Other products are undoubtedly formed. Base treated solutions of ribulose-P2 have a characteristic butter-like odor, similar to the odor of 2,3-butanedione. Therefore, volatile non-phosphorylated compounds must also be present. Evidence for a phosphate elimination mechanism was afforded by following the time course of degradation of [1-13C]xylulose-P2 at pH 11 and 40°C (Figure 10). The starting compound was 90% enriched at C-1. As the reaction proceeded, the area under the doublet centered about 69.4 ppm decreased as methyl resonances at 24.0 and 23.4 ppm increased, reaching their maximum values when no further xylulose-Pz remained. It is concluded that two major products were formed in the alkaline degradation of xylulose-Pz and that these products contain the C-1 carbon of xylulose-Pz as methyl carbons. The methyl resonances are assuned to arise from two major degradation products of the proposed diketo compound via the following mechanism (Chart 3). Similar mechanisms for non-phosphorylated sugars have been proposed (Nef, 1910). Resonances due to the proposed diketo compound (A) cannot be observed with certainty in the experiment of Figure 10, indicating either that this compound is very unstable and rapidly undergoes 86 .Na-omo_=FXXmumH-Hg e_ m=__a:ou ma uu a op mac m_ zag ¢.¢o uzonm cmcmucmo um_a:ou on» .5... 83% 2: meme 24.2 E 2:5 3-: z: 2E: a: c: .2: om a_muos_xocqao mo: maimmo_:_xx mo :o_uocu:oocoo on» .uooe new fifi In no Naummopz—Axmumanau mo :owumvmgmmu we mmgzoo ms_p .oH mzzofim ‘lé—éE set; ? eél CH20P03 CH20P0§ TH20P0= I OH‘ | DH‘ 3 f=0 \ fi-O' \ =0 \ <——_‘— H-T-OH f-O-H Ho-e-H H-f-OH H-f-O-H H- -0H 0 ‘ CH20P0; CHZOPD; (sz (fH3 0 0 1 II II f-OH f=0 + 0H’ \~0 /$ 5 '0 f=o I =0 HO-f COH 3 H3c-T-0H H-f-OH H- -f-0H H- -f-0H H-(i-OH CH20P03 CH20P03 CH20P03 CH20P03 A B C A: 1- -deoxy-D-glycero 2, 3- -pentodiulose-5- -P B: 2- C-methyl -D-threo-tetronic acid-4- P C: 2- wC-methyl-D ethhro -tetronic acid-4-P (Chart 3) a benzylic acid type rearrangement (with subsequent line broadening of the NMR signal) to the compounds (8) and (C), or that the concentration of [1-13C]xylulose-P2 used in this experiment was not sufficient to allow detection of the small amount of this compound that would be expected to be in solution under the basic conditions employed. Since the 2,3-enediol of ribulose-P2 is a proposed intermediate in this mechanism, it would be expected that this compound would also be in equilibrium with xylulose-PZ. The presence of xylulose-Pz 89 in preparations of ribulose-P2 was demonstrated by the following experiments. when substrate amounts of xylulose-PZ were incubated with a large excess of ribulose-P2 carboxylase, significant rates of carboxylation were observed, as if ribulose-P2 were present. The radioactive product of the carboxylation reaction was determined to be glycerate-3-P by descending paper chromatography. That this product was not formed from a slow carboxylation of xylulose-Pz was demon- strated by the observation that product corresponding to only less than 6% of the total xylulose-Pz present was formed regardless of whether the xylulose-Pz solution was allowed to react with the enzyme for 45 min or 90 min. Different batches of xylulose-Pz gave different amounts of glycerate-3-P, but for a given batch, the 45 min and 90 min time points were identical. Therefore, xylulose-Pz is not a sub- strate for ribulose-P2 carboxylase. Furthermore, gas chromatographic analysis corroborated the presence of xylulose-PZ in ribulose-P2 solutions and ribulose-P2 in solutions of xylulose-Pz. Ribulose-Pz and xylulose-PZ were analyzed as their perace- tylated pentitols by gas-liquid chromatography (see Materials and Methods). Ribulose-Pz solutions gave rise to peaks with retention times corresponding to ribitol and arabinitol pentaacetate in approximately equimolar amounts. A small peak (1% of the total peaks observed) corresponding to xylitol pentaacetate was also found along with two more highly retained, unidentified peaks. In like fashion, xylulose-Pz solutions gave rise to two major peaks corresponding to xylitol and arabinitol pentaacetate, a small peak corresponding 90 to ribitol pentaacetate, and a larger amount of the two more highly retained compounds. Therefore, solutions of ribulose-P2 contain a small amount of xylulose-Pz, and solutions of xylulose-Pz contain a small amount of ribulose-P2. The more highly retained compounds may be the acetate derivatives of the branched chain acids that result from the benzylic acid re-arrangement of the proposed dicarbonyl intermediate. Since xylulose-PZ is known to be a potent inhibitor of ribulose-P2 carboxylase (McCurry & Tolbert, 1977), and since a-dicarbonyl compounds such as 2,3-butanedione and phenylglyoxal inhibit the enzyme by binding at arginyl residues at the active site (Paech et al, 1977; Lawlis & McFadden, 1978), the effect of base treatment of ribulose-P2 solutions on enzyme activity was tested. Inhibition of Ribulose-Pp Carboxylase/Oxygenase. Treatment of ribulose-P2 solutions at 30°C and pH 11, followed by the addition of an aliquot of the neutralized ribulose-P2 solution to the enzmne, resulted in enzyme inhibition (Figure 11). This inhibition increased with the time of base treatment to a maximum of 30% residual enzyme activity. Prolonged base treatment or more elevated temperatures caused more rapid substrate loss and orthophosphate appearance but no further inhibition. The increase in inorganic phosphate correlated with the loss of ribulose-P2. These observations may be rationalized by postulating the produc- tion of an unstable enzmne inhibitor from ribulose-P2 (i.e. the diketo phosphate). That is, during the early minutes of base treatment, the inhibitor concentration rose to its maximum value, and then it degraded at a rate approaching its rate of formation so that no 91 FIGURE 11. Effect of incubation of ribulose-P2 at high pH on the time course of the ribulose-P carboxylase reaction. Ribulose-PZ was incubated at pH 11 and 30 for 1,5,10, and 20 min as indicated. The pH was then adjusted to 8.0, and a 0.3 mL aliquot of the ribulose-P2 solution was added to 7.2 mL of assay buffer containing enzyme and NaH[14C]03 so tnat the final concentrations of enzyme and bicarbonate were 80 ug/mL and 20 mM, respectively. Two progress curves were run (o-—-—o). Ten aliquots were withdrawn at indicated times and added to 0.2 mL of 2 N HCl to stop the reaction. The arrows indicate initiation of the second progress curve by addition of a similar amount of non-base treated ribulose-P2. A corresponding volume was removed prior to addition of ribulose-P2 in the second run. Thus, the zero point is 4% lower and a maximum of 96% of the first initial rate in the second progress curve would represent no inhibition. Another solution of ribulose-P2 was incubated for 40 and 160 min at pH 11 and 403C and used for the same experiment (A-—-A). When ribulose-P2 was generated from ribose 5-P enzymatically and used directly, the initial rate in the second progress curve reached 88% of the initial rate in the first progress curve, which because of the dilution, was over 90% of the maximum expected GI---ID. .b umol ”C02 fixed /mg proiein 92 I60 min .— )- .- I l u l L L II I 2 6 7 8 9 )0 ll time (min) 93 further increase in inhibition was noted. In fact, after about 5 h of incubation at pH 11 and 40°C, ribulose-P2 solutions demonstrated no inhibition of the enzyme when assayed in an identical manner. By performing the "two-story“ experiment of Figure 11 at different concentrations of non-base treated ribulose-P2, different amounts of inhibition in the second substrate utilization curve were observed (Figure 12). Doubling the amount of ribulose-P2 essentially doubled the amount of inhibition. Furthermore, ribulose-P2 solutions essentially free of inhibitory effects were obtained in a coupled assay in which ribulose-P2 was generated, inugitu, from ribose-S-P by the action of phOSphoriboisomerase, phosphoribulokinase, and ATP, and assayed directly. In this latter case, there was little effect from inhibitor accumulation excepting the small effect that would be expected from ribulose-P2 degradation even under the mild conditions used for the enzymatic assay. (see I---I in Figure 11). These data allow the following conclusions. Ribulose-PZ solutions give rise to inhibition of ribulose-P2 carboxylase via the production of inhibitors from ribulose-P2. These inhibitors may be produced during preparation, storage, or base treatment of ribulose-P2. Xylulose-Pz is present in preparations of ribulose-P2 and arises from non-enzymatic epimerization. Another inhibitor, a diketo compound, is also present. The rapid formation of inhibitors of ribulose-P2 carboxylase/ oxygenase in solutions of ribulose-P2 is important with respect to ribulose-P2 carboxylase assays and kinetic studies. Recognition of these inhibitors may also be the key to answering some as yet unsolved 94 .5 6* . 2 O a B U? E °. ; 4_ ‘ (as/1‘ 3‘, c “2 ~ (64%). N 8 I 2» , ‘ _ ‘5 E :s b 4 l 22 3 8 59 IC) ll l2 time (min) FIGURE 12. Progress curves of ribulose-P2 carboxylase reactions with (A) 0.5, (B) 0.25, and (C) 0.125 mM ribulose-P2. The experiment was run as described in the legend to Figure 11, except that the ribulose-P2 was not preincubated at high pH. The values in parentheses represent the observed percentage of the initial activity that is observed in the initial velocity of the second progress curve. 95 questions concerning the enzyme in 21129 The average concentration of ribulose-P2 in chloroplasts is of the same order of magnitude, but less than the total concentration of binding sites of ribulose-P2 carboxylase (Jensen & Bahr, 1977). Although conditions for chemical epimerization or B-elimination of ribulose-P2 could easily occur in the alkaline chloroplasts, for example during the day when temperatures may reach 403C or more, the presence of a large amount of ribulose-P2 carboxylase would bind all the ribulose-P2 as soon as it is formed by phosphoribulokinase, as the dissociation constant of the enzyme-ribu- lose-P2 complex is only 1 uM (Hishnick et al., 1970). In this manner, degradation of ribulose-P2.yn vivo may be prevented. CONCLUDING DISCUSSION Data have been presented regarding the use of specifically 13C-enriched carbohydrates in spectrosc0pic studies of their solu- tion behavior. Although the syntheses described in Chapter 1 were directed toward the preparation of the 13C-enriched compounds, modification of the synthesis has allowed the preparation of derivatives containing other stable and radioactive isotopes of C, H, and 0 (Serianni, 1980). These derivatives may well find practical use as analytical and clinical reagents. The specific enrichment with 13C greatly enhances the sensi- tivity of the various NMR spectroscopic methods, and allows the easy interpretation of the resulting spectra. In particular, saturation transfer spectroscopy has proven to be very useful in the study of the reactions of furanose phOSphates. It should prove possible to use this method with other specifically enriched compounds at nominal concentra- tions. A wide variety of potential applications may be easily envisioned. These applications could range from studies of metabolic rates to careful determinations of the mechanisms of Various carbohy- drate reactions. The importance of the tautomerization reaction of enzyme sub- strates is widely appreciated, and further work on this aspect of the chemistry of the sugar phosphates is anticipated. Of particular impor- tance is the determination of whether the rates of tautomerization observed in vitro are the same as the in vivo rates, and whether these 96 97 rates are of physiological importance in determining overall metabolic rates. Since sugar phosphates are common analytical reagents and metabo- lites, a detailed knowledge of their behavior is desirable. In Chapter 3, it was shown how ribulose-P2 degraded and how this degradation affected the enzyme ribulose-P2 carboxylase/oxygenase. This is just one example of the many types of reactions that may be anticipated for these reactive compounds. Unfortunately, the high reactivity of the sugar phosphates has often been only poorly appreciated. It is hoped that the present studies will provide a practical basis for the better understanding of their complex behavior. PART II RIBULOSE BISPHOSPHATE CARBOXYLASE/OXYGENASE: CATALYSIS AND ACTIVATION INTRODUCTION Photosynthetic organisms increase their mass by fixation of carbon dioxide into more highly reduced forms of carbon. In addition, Narburg showed in 1920 that oxygen inhibited carbon dioxide fixation in Chlorella. This inhibition of carbon dioxide fixation by oxygen has since been confirmed in many higher plants as well. In the mid-1950's it was discovered that the primary reaction of photosynthetic carbon dioxide fixation was catalyzed by the enzyme ribulose-P2 carboxylase. The reaction is the condensation of carbon dioxide with a sugar phos- phate, ribulose-P2, to yield two molecules of glycerate-3-P. However, it was not until the early 1970's that the laboratories of 0gren and Tolbert demonstrated that ribulose-P2 carboxylase also catalyzed the addition of oxygen to ribulose-P2 to yield a molecule of glycerate-B-P and a molecule of glycolate-Z-P. The finding that oxygen and carbon dioxide were competitive substrates for the same enzyme immediately linked the oxygenase activity of the enzyme with the oxygen dependent inhibition of net carbon dioxide fixation. Other enzymes have been discovered that cause the release of carbon dioxide from the product of the oxygenase reaction, glycolate-Z-P. This overall process, the light dependent uptake of oxygen and release of carbon dioxide, has been termed photorespiration. The relative fluxes of carbon through the photorespiratory and photosynthetic pathways are such that a large fraction of the carbon dioxide that is fixed during photosynthesis may be released by photorespiration. Thus, 98 99 photorespiration appears to cause the release of the carbon dioxide that was previously fixed at the expense of a large input of energy. These observations, together with the finding that there exists a class of plants, the C-4 plants, that flourish while exhibiting very little photorespiration, have led many investigators to speculate that photorespiration is a wasteful process and should be avoided. The dual function of ribulose-P2 carboxylase/oxygenase consti- tutes the metabolic junction between photosynthesis and photorespira- tion. As such, and for the reasons cited above, many investigators have attempted to characterize the molecular and kinetic properties of this enzyme with the hope that a detailed knowledge of the enzyme will allow the manipulation of the ratio of the carboxylase and oxygenase activities. The approach taken in the following chapters has been to study the reaction mechanism of the enzyme. Chapter l recounts experiments concerning the stereochemistry of the carboxylase reaction. Chapter 2 is a description of the mode of action of various effectors that modulate the enzyme's activities. LITERATURE REVIEW The literature concerning ribulose-P2 carboxylase/oxygenase is vast, and contains well over 1000 articles. Excellent reviews exist (Siegel et al., 1972; Jensen & Bahr, 1977), and reports from a recent symposium on the enzyme have been published (Siegelman & Hind, 1978). This review will focus on the literature that is pertinent to the understanding of the reaction mechanism of ribulose-P2 carboxy- lase/oxygenase. Physical Properties of Ribulose-Pz Carboxylase/Oxygenase. Ribulose-P2 carboxylase/oxygenase from higher plant chloroplasts is a 16 subunit protein with a molecular weight of 560,000 (Kawashima & Hildman, 1970). It is composed of 8 large subunits (56,000) and 8 small subunits (12,000-14,000) (Rutner & Lane, 1967). It is highly soluble and has been estimated to exist in chloroplasts at a concentration of 0.4-0.5 mM (ca. 250 mg/mL), thereby constituting an exceedingly large fraction of the soluble chloroplast protein (Jensen & Bahr, 1977). The enzyme is roughly spherical in shape as determined by ultracentrifugation (Paulsen & Lane, 1966) and electron microscopy (Trown, 1965). The easily crystallized enzyme from tobacco (Nicotiana tabacum) has been studied by X-ray diffraction analysis, and is characterized by a four-fold axis of symmetry and a plane of symmetry parallel to the four-fold axis (04 symmetry group) (Eisenberg et al., 1978). The catalytic sites for ribulose-P2 reside on the large subunit. Therefore, there are_8 binding sites for substrate molecules 100 101 (Hishnick et al., 1970). No clear function for the small subunit has been demonstrated. The enzyme contains a large number of cysteine molecules. Titration of the -SH groups with p-chloromercuribenzoate revealed that some 90 titratable cysteine groups were present per molecule of enzyme (Sugiyama et al., 1968). Calculations based on the ratio of cystine to total amino acids using a molecular weight of 560,000 also indicated the presence of 90 cysteine groups. Therefore, it appears that the cysteine groups are not involved in disulfide bonds in the native enzyme. Perhaps as a result, the enzyme is stabilized by sulfhydryl reagents such as dithiothreitol (N.P. Hall, personal communication). In view of the fact that the enzyme catalyzes an oxygenase reaction, several groups have attempted to determine whether the enzyme contains any metals or co-factors normally associated with oxygenases. One report indicated the presence of I g-atom of copper per mol of enzyme (wishnick et al., 1969), although subsequent studies could not confirm the presence of copper or any other transition metal in a reasonable stoichiometry (Z_1 g-atom metal/mol protein) (Chollet et al., 1975; Lorimer et al., 1973). A recent report that the carboxylase and oxygenase activities of the enzyme were due to two different enzymes, and that the oxygenase enzyme contained copper (Branden, 1978) could not be confirmed by a number of laboratories (McCurry et al., 1978; J.T. Bahr, R. Chollet, unpublished observations). McCurry (1979) presented preliminary evidence that the spinach enzyme contains tightly bound iron to the extent of 2 mol iron/mol enzyme. However, since the level of iron could not be correlated with the amount of oxygenase 102 activity, the possibility remains that this iron was adventitiously bound to the enzyme and is without catalytic function. Finally, attempts to detect the presence of a flavin or other organic cofactors in preparations of the enzyme have been uniformly unsuccessful, even though a wide variety of analytical and physical techniques were used (Chollet et al., 1975; R. Gee & R.M. Mulligan, unpublished observations). The preponderance of data therefore indicates that the enzyme does not contain a catalytically essential transition metal or organic cofactor. Some of the physical and kinetic parameters of ribulose-P2 carboxylase/oxygenase are summarized below (Table 1). Enzyme Activation. Carbon dioxide and a divalent metal ion (usually MgZ+) are essential activators of ribulose-P2 carboxy- lase/oxygenase. Therefore, C02 plays a dual role in the enzymatic mechanism, serving both as an activator of both activities and as a substrate for the carboxylase reaction. The role of C02 as an essen- tial activator was surmised as early as 1963 by Pon et al. (1963) although this phenomenon was not widely appreciated until 1974 when it was shown that the enzyme from freshly ruptured chloroplasts had a higher activity (and lower Km (002)) than the purified enzyme (Bahr & Jensen, 1974a,b). The loss in activity upon purification and storage was correlated with a decrease in [C02] during these processes. This observation set the stage for a more complete description of the activation process (Lorimer et al., 1976; Badger & Lorimer, 1976). These workers showed that C02 reacted slowly and reversibly with a "distinctly alkaline" amino acid residue; this was followed by a rapid reaction of Mg2+ with the enzyme-002 complex. Physical studies 103 Table 1: Summary of Physical and Kinetic Parameters of Ribulose-Pz Carboxylase/oxygenase from Spinacha Holoenzyme: Mw 560,000 Subunits: 8 large MW 56,000 8 small MH 12-14,000 Substrates: ribulose-P2, C02, 02 Essential Cofactor: Mg2+ (or other divalent cations) Reactions: 1. Carboxylase ribulose-P2 + C02 + H20= 2 x (glycerate-3-P + HT) 2. Oxygenase ribulose-P2 + 02=glycerate-3-P + glycolate-2-P + 2H+ Activation: C02 and Mg2+ (or other divalent cations) are required to active both enzyme activities pH Optimum: 8.2-8.6 Kinetic Contentsb: 1. Carboxylase Km (co ) = 10-15 "M Km (ri ulose-PZ) = 20-25 uM kcat (Vm x protomer MH) = 3.5 5‘1 2. Oxygenase Km (0 ) = 0.4 mM Km (r1bulose-P2) = 25-35uM kcat (Vm x protomer MW) 0.3 5'1 aSee references in text. bAt 25-30°c, pH 8. had confirmed that prior formation of the enzyme-C02 complex was necessary for tight metal binding (Miziorko & Mildvan, 1974). The evidence suggested that the enzyme was activated via the following mechanism (Chart 1). 104 CO H Mg2+ 1+ . E-lys-NHZ ‘1”:3‘3. E-lys-N-g -——‘ E-lys-n-g )1ng (slow) \0 (fast) ‘0 (active) Chart 1 That carbamate formation occurs has recently been confirmed by the isolation of a specific lysine-carbamate complex (stabilized by methylation with diazomethane) from proteolytic digests of the activated enzyme (Lorimer, personal communication). Furthermore, it has been shown that the C02 molecule involved in carbamate formation is not the same molecule that is condensed with ribulose-P2 in the catalytic reaction (Miziorko, 1979; Lorimer, 1979). The dual effect of C02 afforded an explanation for the previously observed sigmoidal activity responses to varying C02 concentrations. It has therefore become common practice to differentiate between activation by C02 and catalysis of C02 fixation by preincubating the enzyme with C02 and Mg2+ prior to utilizing the activated enzyme for kinetic studies. When this is done, hyperbolic kinetics are observed, and kinetic analyses are more easily interpreted. A summary of the activation process is given below. 1. Activation entails the slow, readily reversible formation of an enzyme-carbamate-Mg2+ species. As such, the position of equilibrium depends on [HT], [C02], and [M92+]. 2. The kinetics of activation are identical for both carboxylase and oxygenase activities. 105 3. The molecule of C02 that is involved in carbamate formation is distinct from the molecule of C02 that is ultimately incorporated into the carboxylase reaction product, glycerate-B-P. Effectors of Enzyme Activation. The binding constants for C02 and Mg2+ in the activation process are such that the enzyme must be preincubated with up to 300 pM C02 and 20 mM Mg2+ to achieve full activity. Since these concentrations are never present in the chloroplast, a number of investigators have tested other possible activators of the enzyme. Other effectors were found, including gluconate-6-P, fructose-1,6-P2, NADPH, and glycerate-3-P (Buchanan & Schurmann, 1973a,b; Chu & Bassham, 1974,1975). These investigators showed that preincubation of the enzyme with the various sugar phosphates and physiological concentrations of C02 and Mg2+ afforded higher enzyme activities than when the enzyme was preincubated in the absence of the sugar phosphates. Furthermore, an order of addition effect was demonstrated. That is, if sugar phosphate was added simultaneously with the substrate, ribulose-P2, inhibition of enzyme activity was found rather than stimulation. A variety of allosteric models have been postulated to explain these phenomena. One model requires no less than five distinct enzyme sites for interaction with the effectors, four allosteric sites and one catalytic site (Chu & Bassham, 1975). Using this model, it was claimed that these effectors may be of physiological significance in modulating the activity of ribulose-P2 carboxylase/oxygenase lg 1139. It is the purpose of Chapter 2 of this presentation to describe experiments concerning the mode of action of these effectors. The results to be described are consistent with a much simpler model of 106 effector action, and cast doubt on the physiological significance of the effectors in modulating the activity of ribulose-P2 carboxylase/oxygenase. I Mechanism of Carboxylation. An outline for the chemical mechanism of the carboxylation reaction was predicted by Calvin (1954) even before the discovery of the enzyme. This outline has since been extended and clarified by a number of workers (Scheme 1). The initial step is the enolization of ribulose-P2 (1) to form 2, which is attacked by 002 to form a 2-C-carboxy-3-keto-pentitol bisphOSphate, 3. Addition of H20 across the bond at C-2 and C-3 of §_yields two molecules of D-glycerate-B-P (Fiedler et al., 1967; Cooper et al., 1969; Mullhofer‘& Rose, 1965; Pierce et al., 1980; Weissbach et al., 1956; Jakoby et al., 1956). In addition, the oxygen atoms at C-2 and C-3 of ribulose-P2 are retained in the products of the carboxylation reaction, ruling out the intermediacy of eneamine or dithioacetal derivatives in the reaction (Sue& Knowles, 1978; Lorimer, 1978). Attempts have been made to synthesize the B-keto acid intermediate _3_ (Si egel & Lane, 1973) , and quenching from the steady state of the carboxylase reaction gives a compound with the expected properties of 3. (Sjodin & Vestermark, 1973). The finding that a stable analogue of 3, carboxypentitol-Pz, is a competitive inhibitor with respect to ribulose-P2 has been considered proof for the existence of interme- diate §_in the reaction (Wishnick et al., 1970; Siegel & Lane, 1972, 1973). These studies utilized a mixture of carboxyribitol-Pz and carboxyarabinitol-PZ that resulted from cyanide addition to ribu- lose-P2. Although the studies were performed prior to 1976 when the requirement for preincubation with Mg2+ and C02 for maximal Scheme 1 CH 0P0. . 2 3 - H CIO _—————A 1 <---:-- H-(IZ-OH + H H-(IZ-OH CH CPD. 2 3 1 Scheme 2 '_ IN :: o—c'S—o=n—o I 107 CHZOPO HO-C-H ,H 0 C02 _% + $0. “ 2 Hf—OH CH opo' 2 3 _1 CH 090 _| 2 3 C-OH 1'0 H-(IJ-OH CH 0P0 CH 0P0- : a 3 Co 2 1 H20 T _ % $02 + H H-C-OH CH OPO' 108 enzyme activity was clearly established (Lorimer et al., 1976; Badger & Lorimer, 1976), carboxypentitol-Pz was shown to require Mgz+ for maximal inhibition. The enzyme-Mng-carboxypentitol-P2 complex was reported to have Kd < 10-8 M (Siegel & Lane, 1972). These results were suggested to imply a role for Mg2+ in the stabilization of intermediate 3. Studies with [3-3H]ribulose-P2 showed that the tritiated substrate reacted at only 20% of the rate of the unlabeled substrate (Fiedler et al., 1967). Furthermore, following reaction with C02, 98% of the 3H was associated with the medium, indicating that H-3 is not retained in the products. The substantial isotope effect led these and other authors to conclude that enolization may well be rate limiting for the overall reaction. However, the actual situation may not be so straightforward since Arrhenius plots of the carboxylation reaction are nonlinear (Bjorkman & Pearcy, 1971). This non-linearity may reflect the existence of different rate limiting steps at different temperatures, which implies that enolization is not cleanly rate limiting. The manner in which the enzyme controls the stereochemical course of the reaction and the function of Mgz+ in the reaction mechanism are the subjects of Chapter 1 of this presentation. Mechanism of Oxygenation. The most widely accepted mechanism for the oxygenase reaction is one modeled after the carboxylase reaction (Scheme 2). The mechanism is based on the following observations. Studies with [180]2 and H2[180] demonstrated that one oxygen atom from molecular oxygen was incorporated into the carboxyl group of glycolate-P and that one oxygen atom from water was 109 incorporated into the carboxyl group of glycerate-3-P (Lorimer et al., 1973). The observation that oxygen from water was also incorporated into the carboxyl group of glycolate-P was rationalized by proposing that an oxygen exchange from H2[180] into the carbonyl oxygen of ribulose-P2 occurred prior to the oxidative splitting of ribulose-P2 between C-2 and C-3. This proposal has been verified through studies using ribulose-P2 that was specifically enriched with stable isotopes (Pierce et al., 1980). In these studies, 130 from [2-13C1ribulose-P2 was found in the products as [1-13Cngycolate-P, and 180 from [2-1801ribulose-P2 was found in the products as a carboxyl oxygen of [1-180]glycolate-P. Therefore, earlier mechanisms requiring the obligatory loss of 0-2 of ribulose-P2 (Wildner, 1976) may be discounted. There has been a report that the superoxide anion is involved in the oxygenase reaction (Bhagwat & Sane, 1978). This conclusion was based on the observation that bovine erythrocyte superoxide dismutase inhibited the oxygenase activity. The origin of the superoxide anion was unclear. However, other workers have been unable to demonstrate inhibition of the oxygenase activity by bovine superoxide dismutase (Lorimer et al., 1973; McCurry, 1980) or copper penicillamine, a small compound with superoxide dismutase activity (G.H. Lorimer, unpublished observations). Therefore, the intermediacy of the superoxide anion must remain questionable. As the mechanism of the oxygenase reaction is written (Scheme 2) there is no involvement of either an organic cofactor or a transition metal cofactor in the reaction. However, the lack of involvement of one of these cofactors represents quite anomalous behavior for an 110 oxygenase. All other oxygenases that have been studied require at least one of these cofactors or a substrate capable of forming a highly resonance stabilized radical (Hamilton, 1974). These seemingly strict requirements are thought to arise from the fact that 02 exists primarily in the triplet ground state, being separated from the lowest accessible singlet state by an energy barrier of some 22 kcal/mol. The direct reaction of a triplet molecule (02) with a singlet mole- cule (ribulose-P2) to give singlet products (H20, glycolate-P, and glycerate-3-P) is exceedingly unfavorable since this reaction would require that spin inversion of the diradical 02 take place during the time of the Chemical reaction (one molecular vibration, or ca. 10‘135). Since spin inversions are usually much slower processes (lifetimes of ca. 1-10’95), the rate of the direct reaction is expected to be vanishingly small (Hamilton, 1974). Alternatively, 02 may react with a singlet molecule to yield a triplet intermediate which then breaks down to singlet products. However, the large endothermi- city of this process (in the absence of resonance stabilization) argues against its occurence under the conditions of normal enzymatic pro- cesses. These limitations may be avoided by the formation of interme- diate complexes of 02 with transition metals or organic cofactors. These complexes are not properly viewed as triplets since electron delocalization through the metal 3d orbitals or through the conjugated carbon bond system of the organic cofactor (e.g., flavin or quinone) affords a mechanism by which the original electron spins may become uncorrelated. Thus, the spin restrictions are avoided, and the com- plexes can react directly with a substrate molecule or break down to yield an oxidized or reduced species of oxygen which then reacts with 111 the substrate molecule. In either case, the intermediacy of the cofactor has allowed the efficient oxidation of a substrate. As stated earlier, no such cofactors have been found to be cata- lytically associated with ribulose-P2 carboxylase/oxygenase. However, the possibility remains that contaminating metals in the assay system account for the very low rates of oxygenation that are observed (catalytic rate constant ca. 0.35‘1). Alternatively, this very low rate may reflect the very slow direct radical reaction of 02 with ribulose-P2 to form a (poorly) stabilized hydroperoxide intermediate. A very careful study with highly purified buffers and other reaction components will be required to distinguish between these possibilities. If the latter explanation proves to be correct, then ribulose-P2 carboxylase/oxygenase would be a most noticeable exception to the current dogma concerning oxygenase enzymes. 0n the Question of Whether Ribulose-P7 Carboxylation and Oxygenation are Catalyzed by the Same Enzyme. The very low catalytic rate constant for the oxygenase reaction and the apparent absence of transition metal or organic cofactors in oxygenase preparations have led some investigators to speculate that the carboxylase and oxygenase activities in ribulose-P2 carboxylase/oxygenase preparations are due to separate enzymes. This intriguing hypothesis has been widely debated by workers studying this enzyme. The debate centers around the proposition that it is certainly possible that enzyme preparations could contain a contaminating oxygenase at the level of say, 0.1%. With a nominal specific activity and molecular weight, this contaminant could indeed have enough activity to account for the observed specific rate of oxygen uptake in preparations of ribulose-P2 carboxylase/ 112 oxygenase. Therefore, it was with great interest that investigators received the report of Branden (1978) which stated that the two activities could be separated by gel filtration and that the enzyme associated with the oxygenase activity contained copper. However, these results could not be confirmed in a number of laboratories (McCurry et al., 1978; Chollet & Bahr, unpublished observations). The reasons for the disagreement in results remain unclear. However, a large body of data suggests that the activities are, in fact, due to one enzyme. Arguments and observations in support of the one enzyme model include the following. 1. Wherever it is sought, ribulose-P2 oxygenase activity has been found in every enzyme preparation that catalyzes ribulose-P2 carboxylation, regardless of the source of the enzyme (Siegelman & Hind, 1978). 2. For every preparation from a given source, the ratio of carboxylase/oxygenase activities is invariant, excepting the changes in ratio that occur upon changes in temperature (the two reactions have different activation energies (Badger & Andrews, 1974)) and changes in the divalent metal ion used in activation (Wildner & Henkel, 1978). All inhibitors of the carboxylase activity inhibit the oxygenase activity to the same extent. 3. The substrates, C02 and 02, appear to demonstrate competitive behavior in steady state kinetic analyses (Laing et al., 1974; Badger & Andrews, 1974). 4. The kinetics of activation are identical for both activities, depending specifically on [HT], [C02], and [MgZ+] (Lorimer et al., 1976; Badger & Lorimer, 1976). 113 Perhaps the strongest argument in support of the one enzyme model is the last one. That is, it is difficult to imagine the coincidence in which two different enzymes could be activated at identical rates and with identical specificity. The only apparent situation for which this might reasonably occur is if the oxygenase enzyme used as its substrate a product of the carboxylase reaction. This is precisely the situation that has been recently reported to occur (Branden et al., 1980a,b). In these reports, evidence was presented to demonstrate that carboxylase preparations produce L-glycerate-B-P upon reaction of C02 and ribulose-P2 in anaerobic solution. Upon addition of oxygen to the solution, L-glycerate-B-P was said to be converted to glycolate-Z-P with a concomitant uptake of oxygen. Furthermore, an increase in [C02] was said to result in the formation of less L-glycerate-B-P from ribulose-P2 (Branden, personal comunciation). Ostensibly, this would account for the well-known inhibition of the oxygenase activity by C02. Again, these results could not be confirmed with purified spinach enzyme using a large variety of experimental approaches (J. Pierce, unpublished observations; G.H. Lorimer, personal communication). Therefore, since the great preponderance of evidence favors the one enzyme model, any claims to the contrary should be accompanied by the isolation and characterization of another enzyme possessing the qualities of C02 and Mg2+ dependent activation, C02 inhibition, and ribulose-P2 dependent oxygenase activity. Until that time, it is most reasonable and proper to consider the two activities as common functions of a single enzyme, ribulose-P2 carboxylase/oxygenase. 114 Active Site Characterization. A nunber of workers have studied the effects of chemical modification of ribulose-P2 carboxylase/oxy- genase in an attempt to ascertain the identity of active site amino acid residues. A wide variety of specific and rather non-specific reagents have been used. Active site residues that have been identified through the use of general reagents include arginine, as detected by reaction with dicarbonyl compounds such as 2,3-butanedione and phenylglyoxal (Schloss et al., 1978a; Lawlis & McFadden, 1978); tyrosine, as detected by reaction with tetranitromethane (Robison & Tabita, 1979); and cysteine, as detected by reaction with 5,5'-dithiobis-(2-nitrobenzoic acid) (Rabin & Trown, 1964; Trown & Rabin, 1964). The finding that the substrate, ribulose-P2, protected the enzyme from inactivation by these compounds was considered as evidence that the identified amino acids were at the active site. Speculations concerning the possible function of these amino acids in catalysis were also presented, the most widely accepted of which is that arginine promotes substrate binding by electrostatic interaction with the phosphate group of ribulose-P2 (Paech et al., 1978). More specific reagents that have been used include pyridoxal-P for lysine residues (Paech & Tolbert, 1978; Whitman & Tabita, 1978a,b); 3-bromo-1,4-dihydroxybutanone-1,4-P2, which reacted with cysteine and lysine (Norton et al., 1975; Schloss & Hartman, 1977a); and N-bromoacetylethanolamine-P, which also reacted with cysteine and lysine (Schloss & Hartman, 1977b). Amino acid analyses of the oligopeptides obtained by proteolysis of the modified enzyme were performed and several different sequences were observed (Hartman et 115 al., 1978; Spellman et al., 1979). It was shown that pyridoxal-P reacted with the same lysine group (Spellman et al., 1979) as the affinity labels 3-bromo-1,4-dihydroxybutanone-1,4-P2 (Stringer & Hartman, 1978) and N-bromoacetylethanolamine-P (Schloss et al., 1978b). However, this lysine group is not the same lysine group that is involved in carbamate formation (G.H. Lorimer, personal communication), and its function remains unknown. Hartman's group has also synthesized a number of other potential affinity labels including 51;: and trens: 2,3-epoxy-1,4-dihydroxy- butanediol-1,4-P2, N-bromoacetyldiethanolamine-PZ, N-bromoacetyl-- phosphoserine, phosphoglycolic acid azide, and carboxypentitol-Pz azide. Unfortunately, none of these reagents caused inactivation of the enzyme at a sufficient rate to be useful in active site characterizations (Hartman et al., 1978). Sunmary. The widespread appreciation of the centrality of ribu- lose-P2 carboxylase/oxygenase in photosynthetic and photorespiratory carbon metabolism has engendered a great body of facts and Speculation. Perhaps the most noticeably absent data are those that explain the anomalous oxygenase activity and those concerning the stereochemistry of the carboxylase reaction. A detailed description of these and related phenomena will be required before a full appreciation of the enzyme can be attained. It is fortunate then, that the large number of workers in the field and the unabating interest in the enzyme are factors that augur well for the attainment of this happy situation. CHAPTER ONE INTERACTION OF RIBULOSE BISPHOSPHATE CARBOXYLASE/OXYGENASE WITH TRANSITION STATE ANALOGUES The mechanism of the ribulose-P2 carboxylase reaction has been outlined in the literature review. The reaction generates two molecules of D-glycerate-3-P from ribulose-P2 and C02. One of the glycerate-B-P molecules comes from the top two carbons of ribulose-P2 plus C02 and the other comes from the bottom three carbon atoms of ribulose-P2. This stoichiometry requires that a Chiral carbon center (C-3 of ribulose-P2) is lost to become the achiral carboxyl group of the bottom glycerate-3-P, and a chiral carbon center (C-2 of the top glycerate-B-P) is produced from the achiral C-2 carbon of ribulose-P2. The determination of the point at which the enzyme exerts stereochemical control over the reaction is the subject of this chapter. The approach used was to investigate the interaction of the enzyme with various, stereochemically related analogues of the carboxylated intermediate of the carboxylase reaction (Chart 2). This approach is based on the theory that enzymes lower the activation energy of a reaction by binding more tightly to the transition state of the reaction than to the reactants or products (Wolfenden, 1972). This theory predicts that compounds resembling the transition state structure of the reaction will bind more tightly to the enzyme than those that do not. Earlier studies by this approach utilized a mixture 116 CHZOPO3 OZC-C-OH H-C-OH I H-C-OH I - CH20P03 carboxyribitol-P2 eHzopo3 HOH C-C-OH 2 | H-C-OH I H-C-OH | - CHZOPO3 hydroxymethyl-ribitol-P2 EHZOPOB (H0)-C-CHO I H-C-H I H-C-H I CH OPO3 2 dideoxyaldehydo-pentitol-P2 117 Chart 2 (IZHZOPO3 Ho-c-Cog I H-C-OH H-C-OH I = CH20P03 carboxyarabinitol-P2 CHZOPO3 HO-C-CHZOH I H-C-OH I H-C-OH I = CHZOPO3 hydroxymethyl-arabinitol-P2 CIHZOPO3 I H-C-OH I H-C"' O I CHZOPO3 aldehydo-pentitol-P2 118 of carboxyribitol-Pz and carboxyarabinitol-Pz that resulted from cyanide addition to ribulose-P2. These compounds resemble the proposed intermediate of the carboxylase reaction (see Literature Review), and indeed the mixture was reported to bind to the enzyme with a dissociation constant of less than 10"8 M (Wishnick et al., 1970; Siegel & Lane, 1972,1975). However, the earlier workers did not determine the individual affinities of the two components for the enzyme. In view of our interest in the stereochemistry of the carboxylase reaction and in the use of the carboxypentitol-Pz compounds by a number of workers to probe the active site of the enzyme (Miziorko.& Mildvan, 1974; Ryan 8 Tolbert, 1975; Schloss et al., 1978; Miziorko, 1979), it was desirable to further characterize these compounds and their interaction with the enzyme. Consequently, structural analogues (see Chart 2) of the reaction intermediate in the carboxylation of ribulose-P2 were synthesized and their interactions with ribulose-P2 carboxylase/oxygenase were examined. Correlation of the stereochemistry of these compounds with their effect on the enzyme allowed conclusions regarding the stereochemistry of the carboxylated reaction intermediate and the role of C02 and Mgz+ in the carboxylase reaction. Materials and Methods Compounds and Enzymes. General techniques and the preparation of sugar phosphates have been described earlier (Part I; Chapter 1). 1,5-Dihydroxypentan-Z-one-Pz was prepared by the procedure of Hartman.& Barker (1965). L-glycerate-B-P was prepared by oxidation 119 (Shaffer & Isbell, 1963) of L-glyceraldehyde-3-P which was produced by the action of glycerolkinase and ATP on DL-glyceraldehyde (Serianni, 1980). Ribulose-Pz carboxylase/oxygenase was purified from spinach (Ryan & Tolbert, 1975); the enzyme was stored as frozen pellets at -80°C. Prior to use, the pellets were thawed, and the thawed solution was made to 50 mM in dithiothreitol. To remove the excess reducing agent, the solution was dialyzed against 100 volumes of 50 mM Bicine buffer (pH 8.2) that contained 0.1 mM EDTA. This procedure was required to get reproducibly high specific enzyme activities. Carboxyribitol-P7_[2-C-(Phosphohydroxymethyl)-D-arabinonic Acid- S-P] and Carboxyarabinitol-Pp_[2-C-(Phosphohydroxymethyl-D-ribonic ‘Agid1§:_l. A 50-100 mM solution of ribulose-P2 (with or without 13C and/or 14C enrichment) at pH 8.5 was added to a 0.5 M solu- tion of KCN (with or without 13C and/or 14C enrichment) so that the final ratio of cyanide to ribulose-P2 was 1.1. The resulting nitriles were allowed to hydrolyze to the acid salts at 22°C for 48 h. The solution was treated with excess Dowex 50(HT), filtered, concen- trated to dryness _i_n_ 339113 at 30°C, and desiccated ill 313933 ((0.1 man) at room temperature over MgCl04 for 24 h. The lactones were dissolved in water and quickly adjusted to pH 5.5 with 1 M NaOH. For small amounts of material (<1 mmol of total P), the solution was added to a 42 X 2 cm Dowex 1 (8% cross-linked; 200-400 mesh) column in the chloride form and eluted with a 4-L, linear gradient of 0.0-0.4 M LiCl in 3 mM HCl at a rate of 0.5-0.8 mL/min. For larger amounts of material, the solution was added to a 49 x 3.3 cm Dowex 1(Cl‘) column and eluted with a 6-L, linear gradient of 0.0-0.4 M LiCl in 3 mM HCl at a rate of 0.6-1.0 mL/min. Fractions (15 mL) were collected and 120 assayed for total phOSphate (LeLoir & Cardini, 1957) or radioactivity. Peak fractions were pooled, neutralized to pH 8.0 with I M Li0H, and concentrated in £2222.t° approximately 100 mL. The addition of a threefold molar excess of barium acetate, followed by the addition of ethanol to a final concentration of 50% (v/v), precipitated the bisphosphates as their barium salts. After at least 1 h at ~203C, the precipitate was collected by centrifugation and twice washed with 95% ethanol. The products were dissolved in water by the addition of excess Dowex 50(H+), filtered, adjusted to pH 6.5 with 1 M NaOH, and stored at -20°C until use. Recovery of radioactivity was usually about 90% of that applied to the Dowex 1(Cl‘) column. The compounds were estimated to be at least 95% pure by 13C NMR, phosphate, and gas chromatographic analyses. Prior to use with ribulose-P2 carboxylase/ oxygenase, the compounds were incubated at pH 9.0 for 24 h at room temperature to ensure that no lactone forms were present. Hydroxymethylribitol-P9_[2-C-(Hydroxymethyl)-D-ribitol-P9]_and Hydroxymethylarabinitol-Pp_[2-C-(Hydroxymethyl)-D-arabinitol-Ppl. The Y-lactones of [2'-13C]-carboxyribitol-P21 and [2'-13C]- carboxyarabinitol-Pz were prepared from their salts by Dowex 50(H+) treatment and desiccation as described above. A 20-mL aliquot of a 0.5 M solution of NaBH4 in 0.4 M Na2C03 was added to 50 pmol of the appropriate lactone. The reduction was terminated after 24 h at room temperature by the addition of 2 mL of glacial acetic acid. 1The branched-chain compounds used in this report are named as derivatives of the D-pentitol-Pz compounds and numbered so as to stress their structural relationship to ribulose-P2. The tertiary carbon is designated C-2. The carbon derived from cyanide (i.e., carboxyl, hydroxymethyl, or aldehydo carbon) is designated C-2'. 121 Dowex 50(H+) treatment, evaporation to dryness lg vague, and repeated concentration from anhydrous methanol removed excess borate. The products were purified by column chromatography on Dowex 1(Cl‘) as described above. The hydroxymethyl derivatives eluted at approximately 0.1 M LiCl, followed by the carboxy derivatives; products were col- lected as their barium salts, converted to the sodium salts, and stored at -20°C as described for the carboxypentitol bisphosphates. [2'-13C]- Hydroxymethylribitol-Pz and [2'-13C]hydroxymethylarabinitol-P2 were characterized by 13C NMR analysis. The 13C NMR chemical shifts of the enriched carbons were 62.8 ppm for the ribo derivative and 63.5 ppm for the arabino derivative. The purity of these compounds was estimated to be at least 95% by 13C NMR. Aldehydopentitol-P2f£Equimolar Mixture of 2-C-(Phosphohydroxy- methyl)-D-ribose-S-P and 2-C-(Phosphohydroxymethyl)-D-arabinose-5-P . An equimolar solution of K[13C]N (containing K[14C]N) and ribulose-P2 was kept at pH 7.5 and 4°C for 45 min. The resulting cyanohydrins were reduced on palladiumébarium sulfate (5%) (See Part 1, Chapter 1). Analysis by 13C NMR revealed the presence of glycosy- lamine derivatives after the reduction. The mixture of glycosylamines was converted to the aldoses by incubation at pH 8.4 for 12 h at room temperature. Aldehydopentitol-Pz was purified on a 51 x 2.2 cm Dowex 1(formate) column by using a 4L, linear formate gradient (0.2-1.3 M, pH 6.2). The aldoses eluted at approximately 0.5 M sodium formate. After treatment of the pooled fractions with excess Dowex 50(H+), formic acid was removed by continuous ether extraction for 24 h at 4°C, and the compounds were stored as their sodium salts at -20°C until use. Analysis by 13C NMR revealed that aldehydopentitol-Pz exists in 122 solution as the cyclic, anomeric furanoses (chemical shifts: 102.9 ppm, 43%; 102.2 ppm, 8%; 98.3 ppm, 49%) with no detectable ((4%) gem-diol or free aldehydo forms. The purity of these compounds was estimated to be at least 95% by 13C NMR. Aldehydo-3,4-dideoxypentitol-P2_[(RS-Z-C-(Phosphohydroxy- methyl)-3,4-dideoxypentose-S-P]. An equimolar solution of K[13CJN (containing K[14C]N) and 1,5-dihydroxypentan-Z-one-Pz was kept at pH 8.0 and 22°C for 30 min. The resulting cyanohydrins were converted to the aldoses by catalytic reduction on palladium-barium sulfate (5%) (See Part I, Chapter 1). Aldehydo-3,4-dideoxypentitol-P2 was purified by Dowex 1(Cl‘) chromatography (Byrne & Lardy, 1954); it was isolated as its barium salt and stored at -20°C in the manner described for the carboxypentitol bisphosphates. Analysis by 13C NMR revealed the presence of 70% gem-diol (93.2 ppm) and 30% free aldehyde (207.6 ppm) forms at 303C and pH 7.0. The purity of these compounds was estimated to be at least 90% by 13C NMR. Characterization of Carboxyribitol-P; and Carboxyarabinitol-P . Either carboxyribitol-Pz-y-lactone or carboxyarabinitol-Pz y-lactone was treated with acid phosphatase (potato) at pH 4.5, depro- teinized (Somogyi, 1945), applied to a 20-mL column of Dowex 1(Cl'), and eluted with water. The dephosphorylated y-lactones (carboxypen- titol v-lactones) were concentrated to dryness and desiccated igivggug. for 48 h. The compounds were derivatized in pyridine with N,0-bis(tri- methylsilyl)trifluoroacetamide containing 1% trimethylchlorosilane for gas Chromatography-mass spectrometric analysis. The carboxypentitol Y-lactones were further characterized by 13C NMR analyses and by oxidation with periodate (Ferrier, 1962). Rates of periodate reduction 123 were followed spectrOphotometrically (Dixon & Lipkin, 1954). The infrared spectra of the lithium salts of the carboxypentitol bisphosphates (acid and lactone forms) were obtained in KBr pellets. General Assay Procedures for Ribulose-Pp Carboxylase/Oxygenase. Enzyme at a concentration of 0.4-2 mg/mL was activated with C02 (Lorimer et al., 1976) at 30°C for at least 30 min in assay buffer (0.1 M N,N-bis(2-hydroxyethyl)glycine (Bicine), pH 8.1, 20 mM MgCl2, 1 mM dithiothreitol, and 0.1 mM Nag EDTA) containing 10 mM NaHCO3. Ribulose-PZ carboxylase activity was determined by the radiometric assay (Paulsen & Lane, 1966). Activated enzyme was added to the assay buffer which contained 10 mM NaH[14C]o3 (0.14-1.4 Ci/mol). The reaction was initiated with ribulose-P2. Under these conditions the specific activity of different enzyme preparations varied between 1.5 and 2.2 nmol of C02 fixed per min per mg of protein. Kinetic parameters, vmax and Km, were calculated with a computer program using a nonlinear regression analysis (Wilkinson, 1961). For the determination of inhibition constants (K1), 0.46 mL of an enzyme solution was mixed with 40 uL of a solution containing ribulose-P2 and inhibitor at appropriate concentrations so that the desired final concentrations were achieved. Averages of duplicate measurements are reported. The error in the assay was less than 5%. For experiments with carboxyarabinitol-Pz, the assay time was 15 s. Assay times for other inhibitors were 60 s. Linearity of the assay was confirmed for all assay conditions and assay times reported. Protein concentrations were determined by the method of Bensadoun and Weinstein (1976) or by absorbance at 280 nm with 2%30 = 16.4 (Paulsen & Lane, 1966). 124 FIGURSI Separation of carboxyribitol- -P2 and carboxyarabinitol- P2 and C NMR analysis of the separated compounds. (A) Chromatog- raphy of the reaction products of cyanide addition to 1. 2 mmol of ribulose-P2 on a 49 x 3.3 cm Dowex 1-X8 column in the chloride form with a linear gradient of LiCl (6 L, 0.0-0.4 M) in 3 mM HCl. Peaks 3 and 4 contain, reipectively, 40 and 32% of ghe material applied to the column. )The H-decoupled, 15.08 MHz, 1 C NMR spectra of the compounds (from peaks 3 and 4. Spectra were obtained at 25°C, pH 2.5, with a sweep width of 3000 Hz. Unassigned resonances arise from hydrolysis of the compounds during the NMR analysis. 125 / 0.4 4.0 . A / / I / 2 3 / E _ / . 0.3 .—-. 30 I. / I :5. / I 5 "/ / 4 ' a 2* 2 g 2.0 - / ’ “ 0.2 .— 5 e a 6 1) .L‘; LO . 0.! 5 6 B d b at O C Peak 3 IVJI d a b ef C I Peak 4 I H I IU|I Iéo Iéo I40 I20 160 8'0 6'0 126 Results Purification and Characterization of the Carboxypentitol Bisphosphates. The chromatographic separation of the products of cyanide addition to ribulose-P2 is presented in Figure 1. The compound from peak 1 and the compound from peak 2 appear to be the free acid forms of the compound from peak 4 and the compound from peak 3, respectively. This was demonstrated by the observation that rechromatography of the peak 1 compound gave peaks 1 and 4, while rechromatography of the peak 2 compound gave peaks 2 and 3. Peaks 1 and 2 were not observed if very careful lactonization was performed prior to column chromatography. Alternatively, if the products of cyanide addition to ribulose-P2 were chromatographed without prior acid treatment and desiccation, partial lactonization occurred on the column, peaks 1 and 2 predominated, and a smeared elution profile resulted. When K[14C]N was used, all peaks contained compounds that had a ratio of total phosphate/14C of 2:1. The peak areas in the elution profile indicate that the cyanohydrin synthesis, as applied to ribulose-P2 here, gives two epimeric products in approximately a 1:1 ratio. In addition 130 NMR resonances of equal intensity at 178.1 and 176.7 ppm were observed after lactonization when K[13C]N was used, confirming the equal distribution of products. The infrared spectra of the pentalithium salts of the compounds. _ from peaks 3 and 4 gave absorptions typical of carboxylic acids at 1620 cm'l. After treatment with Dowex 50(H+) and desiccation, infrared absorptions typical of Y-lactones at 1780 cm'1 were 127 Table 2. Properties of the v-Lactones y-arab- peak 3 peak 4 inono- y-ribono- compd compd lactone lactone periodate redg, 330 15 750b 23b tI/Z (min)a: 13C-1H coupling 0.5 2.6 constant, 3 JIHs-Cs-Czecl) (H2)c’f 13C NMR chemical C-2 78.7 75.7 73.2e 70.4e shifts (ppm)d C-3 74.7 68.7 73.8e 70.8e aTime required to reduce 0.5 molgr equiv of periodate at 25°C; E104’] = B -lactone] = 5.4 x 10‘ . From Ferrier (1962). CDetermined at 25°C and pH 7.0 with a sweep width of 1500 Hz. The goupling constants are accurate to within 0.7 Hz. Conditions are given in Figure 1. eFrom H.A. Nunez (unpublished observations). Measurements were made with the dephosphorylated-y-lactones. observed (Barker et al., 1958). Mass spectrometric analyses showed that these compounds were epimers since similar fragmentation patterns were obtained, the only differences being in the relative abundances of the mass fragments (Petersson, 1970). The 13C NMR spectra of the compounds from peaks 3 and 4 show resonances of six carbon atoms (Figure 18). Assignments of the resonances were made by using acids prepared from ribulose-P2 selectively enriched with 13C at C-1 and the patterns of carbon-phosphorus coupling (Lapper et al., 1973; Lapper & Smith, 1973). In addition, by the use of K[13C]N to 128 enrich the carbonyl carbons, it was shown that the epimeric carbon atoms with resonances at approximately 76 and 79 ppm were coupled to the carbonyl atoms by approximately 56 Hz, indicating direct bonding of these atoms. Unlabeled resonances in Figure 1 arise from the free acid forms due to hydrolysis of the y-lactones during the NMR analysis. The only structures consistent with the above data are those for carboxyribitol-P2 and carboxyarabinitol-P2. The absolute stereo- chemistry about C-2 of these compounds was determined by comparing the rates of periodate oxidation of their dephosphorylated y-lactones. Carboxyribitol-Pz forms a y-lactone with the C-2 and C-3 hydroxyl groups in a 25333 configuration, while carboxyarabinitol-P2 forms a y-lactone with the C-2 and C-3 hydroxyl groups in a gj§_configuration. The rate of periodate oxidation of adjacent hydroxyl groups depends on the stereochemistry of the glycol group being oxidized, gisrglycols being oxidized much faster than transyglycols (Ferrier, 1962). The dephospholactone from peak 4 is oxidized much more rapidly than that from peak 3 (Table 2). Production of formaldehyde from the oxidation of the 1,2-glycol accounted for less than 5% of the oxidation products after 0.6 molar equiv of periodate had been reduced. Thus, periodate cleaved predominantly at the 2,3-glycol, and the differences in the rates of oxidation reflect differences in the stereochemistry at this site (Ferrier, 1962). On this basis, the peak 4 compound is tentative- ly identified as carboxyarabinitol-P2 y-lactone. This assignment is supported by analysis of 13C NMR spectra of the lactones synthesized from [1-13C]ribulose-P2. The angle between H-3 and C-1 is different in the two compounds, being close to 0° for carboxyarabinitol y-lactone and approximately 110° for the ribo 129 epimer. The coupling constants for 13C and 1H nuclei separated by three bonds, in this case 3J(H3-C3-C2-C1)’ depends on the angle subtended by the bonds (Schwarz & Perlin, 1972; Perlin et al., 1974), being maximal at 0 and 180° and minimal at 90°. The three-bond coupling constant obtained from the 1H-coupled, 13C NMR spectrum of the dephospholactone from peak 4 was larger than that observed in the peak 3 dephosphopholactone (Table 2), indicating that peak 4 contains the arabino isomer, in agreement with the conclusion drawn from the periodate oxidation experiments. In addition, the correctness of the above assignment is supported by the 13C NMR chemical shifts of the ring carbons of the epimeric lactones. The dependence of chemical shifts on stereochemistry has been demonstrated for the furanose ring system of pentose 5-phosphates (Part 1, Chapter 1) and for the y-lactone ring system of pentono-y-- lactones (H.A. Nunez, unpublished observations). When hydroxyls at C-2 and C-3 are gig, the chemical shifts of these carbon atoms are at a higher field than when the C-2 and C-3 hydroxyls are trans. The chemi- cal shifts for C-2 and C-3 of the peak 4 compound occur at higher field (smaller chemical shift values) than those for C-2 and C-3 of the peak 3 compound (Figure 1; Table 2) in keeping with the stereochemical assignment made above. Although none of the above criteria constitutes proof of the stereochemical assignment, the agreement between all of the chemical and physical analyses supports the identification of the peak 3 com- pound as carboxyribitol-Pz v-lactone and of the peak 4 compound as carboxyarabinitol-P2 v-lactone. Since the 2-C-hydroxymethyl deriva- tives were prepared from the separated lactones, their stereochemistry I30 6 i I I 3 . - E 4_ q 02 .'s 8 . Q E a E - 01» 3 2- a _'> C , 6 I6 l5 [corboxyribiiol - P2 ] u M Io 50 so . I mM.l [rIbulose-PZ] FIGURE 2. (Left) Lineweaver-Burk plot of the inhibition of ribulose-P2 carboxylase activity by carboxyribitol-Pz. The general assay was used with no inhibitor (A) or with 2, 4, 8 or 16 pM concentrations of carboxyribitol-Pg (0). Km (ribulose-P2) = 36 pM. (Right) A plot of the slopes of the Lineweaver-Burk plot vs. the concentration of carboxyribitol-Pz. Ki = 1.5 pH. 131 is also established. Interaction of the Carboxypentitol Bisphosphates with Ribulose-Pp Carboxylase/Oxygenase. The inhibition of ribulose-P2 carboxylase/oxygenase by carboxyribitol-Pz (Figure 2) is strictly competitive with respect to ribulose-P2 with Ki = 1.5 uM (assuming rapid equilibrium between enzyme and carboxyribitol-Pz). The rapid equilibrium assumption was verified by preincubating the activated enzyme with inhibitor for 5 s to 50 min prior to the addition of ribulose-P2. The preincubation was without effect on the subsequent assay. Therefore, the equilibrium is achieved rapidly (within 5 s) with respect to the assay time (1 min). Equilibrium dialysis studies which gave Kd = 1.5 uM for the enzyme-carboxyribitol-Pz complex indicate that there are eight binding sites per 560,000 daltons (data not shown). The binding is not cooperative. The inhibition of ribulose-P2 carboxylase/oxygenase (E) by carboxyarabinitol-P2 (I) is time dependent, as reported by Siegel and Lane (1972) for the epimeric mixture. They proposed a mechanism (eqn 1) based on the observation that enzyme at 4°C bound carboxypentitol-P2 k1 k3 E+I—'=‘EI<._—::EI* (1) E2 k4 tightly but still retained full activity in the standard carboxylase assay. Incubation at 30°C was reported to be required for enzyme inhi- bition. In addition, they observed that the irreversible inhibition of the enzyme by carboxypentitol-P2 at 30°C was second order. None of these results could be obtained with carboxyarabinitol-P2. 132 These findings, together with the known flexibility of the enzyme and its conformational sensitivity to temperature (Chollet & Anderson, 1976, 1977; Wildner & Henkel, 1977), prompted experiments designed to test the proposed mechanism of inhibition directly. The mechanism is conceptually identical with the Michaelis-Menten mechanism for enzymatic catalysis. Therefore, the rate of inhibition by carboxyarabinitol-P2 should demonstrate saturation kinetics. Thus, if k4 is much smaller than the other three rate constants, then, when [carboxyarabinitol-P2] >> [enzyme active sites], the improved steady-state approximation of McDaniel and Smoot (1956) can be applied to yield the integrated rate equation - E [EJt - T‘TCTITI7K;7"' exp("kobsdtI (2) where obsd ‘ Ks + I (3) ... ... erg—£2 I.) The rate of inhibition of the enzyme by carboxyarabinitol-P2 may be and determined by measuring the amount of enzyme activity remaining after preincubation of the enzyme with carboxyarabinitol-P2 for various times and at various inhibitor concentrations. When enzyme activity (V) is proportional to free enzyme concentration, a plot of ln (Vt/vcontrol) vs. time of preincubation should yield straight lines with slopes (k ) varying with inhibitor concentration obsd according to eqn 3 and with extrapolated, ordinal intercepts varying with inhibitor concentration according to the preexponential term in 133 eqn 2. The proportionality of enzyme activity with enzyme concentra- tion will hold only if the El complex is catalytically incompetent. In this case, eqn 3 indicates a hyperbolic relationship between kobsd and [I], and a plot of kobsd vs. [carboxyarabinitol-P2]"1 should yield a straight line with an intercept on the ordinate of 1/k3 and a slope of KS/k3. As shown in Figure 3, the data are consistent with these predictions and indicate a rapid equilibration of the enzyme with carboxyarabinitol-P2, followed by a slow interaction of the enzyme with the inhibitor which has a rate constant k3 = 0.04 5‘1. Additionally, this slow, first order process appears to be accompanied by a conformational change in the enzyme. Siegel and Lane (1972) have reported that upon addition of carboxypentitol-P2 or ribulose-P2 to the enzyme, a difference spectrum of the sugar phos- phate-enzyme complex versus enzyme gives a positive absorption at 288 nm and a negative absorption at 296 nm. This observation was confirmed for carboxyarabinitol-P2. Furthermore, in the presence of MgZT, the formation of the maximum and minimum in the difference spectrum was time dependent for carboxyarabinitol-P2. Upon addition of the inhi- bitor to the enzyme, the magnitude of the maximum at 288 nm and the minimum at 296 nm increased with time in a first order fashion with k = 0.04 5'1 (data not shown). The agreement between the rate of formation of the difference spectrum and the rate of formation of the EI* complex suggests that the two processes are equivalent, and that the formation of the E1* complex involves a conformational change in the enzyme. In the absence of M92+ a similar difference spectrun was observed for the interaction of carboxyarabinitol-P2 with the enzyme, although the magnitudes of the maximum and minimum were 134 FIGURE 3. Kinetics of inhibition of ribulose-P2 carboxylase activity by carboxyarabinitol-P2. A 0.1 mL aliquot of a solution containing carboxyarabinitol-P2 at various concentrations was rapidly mixed with 4.9 mL of assay buffer at 25°C containing activated enzyme and NaH14CO3 (1.4 Ci/mol) so that the final concentratons were 7 ug of enzyme per mL and 10 mM NaH14C03. The final concentration of carboxyarabinitol-P2 was varied between 0.0 and 4.8 pM. At indicated times, 0.48-mL aliquots of these solutions were added to 20 pL of a 25 mM ribulose-P2 solution, and the extent of C02 fixation was ‘determined over a 15-s period. (A) Pseudo-first-order plot. See the text for details. Concentrations of carboxyarabinitol-P2 were 0.3 (o), 0.45 (A), 0.60 (D), 1.20 (I), 2.40 (o), and 4.80 M (+). The lines were drawn from fitting the data by regression analysis. (8) Double-reciprocal plot of the rate constants obtained in (A) with varying concentrations of carboxyarabinitol-P2. k3 = 0.04 5‘1. Ks = 006 ”Mo 135 2K3 I .4() END time (sec) Figure 3A 136 8C) E3 .... 60- O Q) 3 0) .0 — O x. 4o- 20 I 2 3 4 [carboxyarabinitol - Pa] Figure 3B 137 smaller and the formation of the extrema was very fast. The rapid interaction of carboxyarabinitol-P2 with the enzmne was investigated by measuring inhibition when carboxyarabinitol-P2 and ribulose-P2 were added simultaneously to the enzyme. Short assay times (15 s) were used to minimize the effects of the slow, second phase of inhibition (Figure 4). The first phase of inhibition by car- boxyarabinitol-P2 is purely competitive with ribulose-P2 with K; 0.4 pM. This value is in excellent agreement with the value of KS 0.6 pM obtained by using the steady-state assumption (Figure 3; eqn 4) and validates the assumption that the E1 complex is catalytically incompetent. Dissociation of Carboxyarabinitol-P; from Ribulose-Pp Carboxylase/Oxygenase. The rate of exchange of enzyme-bound, radio- active carboxyarabinitol-P2 with added, nonradioactive carboxyarabin- itol-P2 was measured (Table 3) to evaluate k4. In the presence of C02 and MgZ+, only 6-7% of the labeled carboxyarabinitol-P2 dissociated from the enzyme in 5 days. Complete dissociation is effected by protein denaturation with sodium dodecyl sulfate. The close agreement between the values of the thermodynamic constant, K; = kz/kl, and the kinetic constant, KS (eqn 4), indicates that k2 is greater than k3. Since the values of kg and k3 are much greater than the rate of dissociation of carboxyarabinitol-P2 from the enzyme, the rate-limiting process in this dissociation is that process related to k4. When the observed rate of exchange is equated with k4, an upper estimate of k4 from these data is 5 x 10"7 5‘1.2 The equilibrium constant for the second phase of inhibi- tion depicted in eqn 1 is therefore K2 = [EI]/[EI*] = k4/k3 5 1.2 x 138 m5- '3 E QIO- % "0.05- E 5 2 . . . _|> 2 4 6 [carboxyarabinitol-EILLM IO 20 30 | -| —— m [n bulose-Pz] FIGURE 4. (Left) Lineweaver-Burk plot of the inhibition of ribulose-P2 carboxylase activity by carboxyarabinitol-P . The general assay was used with no inhibitor (A) or with 0. , 1.6, 3.2, or 6.4 uM concentrations of carboxyarabinitol-P2 ( ). Km (ribulose-P2) = 20 pM. (Right) A plot of the slopes of the Lineweaver-Burk plot vs. carboxyarabinitol-P2 concentration. 0.4 uM. K1= 139 10'5. The overall binding constant for the interaction of carboxyarabinitol-P2 with ribulose-P2 carboxylase/oxygenase is then 11 M. This value is 105 times less than the K = K'iKZ $10- K; of 1.5 pM determined for carboxyribitol-Pz and also 105 times smaller than the dissociation constant for ribulose-P2 (Wishnick et al., 1970). Effects of Mg2+ and CD; on the Interaction of Carboxyara- binitol-Pp with Ribulose-Pp Carboxylase/Oxygenase. Siegel and Lane (1972) demonstrated that Mg2+ is essential for maximal inhibition of ribulose-P2 carboxylase/oxygenase by carboxypentitol-P2. This result has since been confirmed and clarified in studies which indicate that enzyme active sites, Mng, C02, and carboxypentitol-P2 form a quaternary complex in a 1:1:1:1 stoichiometry (Miziorko, 1979). Since the published data are most consistent with the data given above for carboxyarabinitol-P2, the effect of Mg2+ on the binding of carboxyarabinitol-P2 to the enzyme was investigated. It has been shown that carboxyarabinitol-P2 is bound very tightly to the enzyme (K < 10'11 M) in the presence of MgZ+. In the absence of MgZT, exchange between bound and unbound carboxy- arabinitol-P2 occurs relatively rapidly (Table 3) although the compound binds tightly enough to remain with the protein during a gel filtration. Thus, only the second phase of inhibition by carboxyara- binitol-P2 exhibits a requirement for Mng. It is not known whether carboxyarabinitol-P2 binds to the enzyme in a two-step 2This value is considered an upper estimate. Although the apparent rate of dissociation is first order in [EI*], the enzyme may denature during the time that would be required to determine the rate constant k4 over a greater percentage of the exchange reaction. 140 Table 3. Dissociation of Carboxyarabinitol-Pz from Ribulose-Pz Carboxylase/Oxygenase ' moi of [2'-14-C]- Time carboxyarabinitol-P2 Conditions (h) per mol of enzyme +Mg2+ O 6.3 24 6.2 48 6.0 96 6.0 120 5.8 +NaDodSD4 120 0.0 1 0.8 aRates of dissociation of the enzyme-carboxyarabinitol-P2 complex were determined by following the time course of release of radioactive carboxyarabinitol-P2. Ribulose-P carboxylase/oxygenase (30 mg, 36 nmol) was incubated for 1 h at 25 with 0.9 pmol of [2'- CJCar- boxyarabinitol-Pz (0.81 Ci/mol) in a final volume of 1 mL of assay buffer containing 10 mM NaHC03'with or without 20 mM MgClz. This solution was applied to a 60 x 1 cm Sephadex G-25 column previously equilibrated with the same buffer solution. Elution with buffer solution separated unbound carboxyarabinitol-P2 from the enzyme. Fractions containing protein were pooled, 29 nmol of nonradioactive carboxyarabinitol-P2 was added, and the solution was adjusted so that its final composition was 2 mg/mL enzyme, 50 mM Bicine (pH 8.1), 2 mM dithiothreitol, 10 mM NaHC03, and 2.9 mM carboxyarabinitol-P2 (with or without 20 mM MgClz). The rate of exchange of radioactive carboxyarabinitol-P2 with unlabeled carboxyarabinitol-P2 at 25°C was determined over a 5-day period by passing I-mL aliquots of the incubation solution over the gel filtration column, and assaying for radioactivity and protein (see Materials and Methods). After 5 days, the solution was made 1% in sodium dodecyl sulfate and a 1 mL aliquot was passed over the gel filtration column as before. 141 fashion in the absence of Mg2+, nor can this question be addressed by following enzyme activity since Mg2+ is required for activation of the enzyme. This activation, which also requires the presence of C02, has been recently shown to involve the formation of a carbamate by reaction of C02 with the epsilon amino group of a lysine residue (G.H. Lorimer, personal communication). The fact that 002 is required for tight binding of MgZ+ (Miziorko & Mildvan, 1974) suggests that the negative species created upon carbamate formation is stabilized by complexation with Mng, as shown in Chart 1 (p. 104). Since both activation of the enzyme and binding of carboxyarabini- tol-P2 depend on MgZT, and since the binding of M92+ requires the presence of C02, a correlation between the extent of enzyme activation and the extent of binding of carboxyarabinitol-P2 was sought. I When the enzmne was preincubated with varying concentrations of C02 and M92+ it was found that the tight binding of carboxyarabinitol-P2 depended on C02 and Mg2+ concentrations in the same manner that enzyme activation depended on the concentration of these species (Figures 5A and 58). Therefore, carboxyarabinitol-P2 is tightly bound only to the active ECM form of the enzyme, although it also binds (less tightly) to the E and EC forms. A reasonable explanation for the Mg2+ requirement is that the negatively charged carboxyl group of carboxyarabinitol-P2 interacts with the positively charged metal ion. This proposal is supported by the observation that the 13C NMR resonance of the carboxyl group of carboxyarabinitol-P2 is strongly affected by low concentrations of paramagnetic ions such as Mn2+ which do not affect resonances of 142 FIGURE 5. The effects of C02 and Mgz+ on enzyme activity and carboxyarabinitol-P2 binding. Enzyme (1 mg/mL) was incubated in degassed, 0.05 M Bicine buffer, pH 8.0, with various concentrations of NaHCO3 and MgClz for 15 min at 22°C in air tight vials (total volume = 1 mL). A 20 uL aliquot of [2'-14C]carboxyarabinitol-Pz (0.81 Ci/mol) containing a 1.5 fold excess of the inhibitor over potential active sites was added, and the solution was kept for 45 min; a 100 fold excess of non-radioaizive carboxyarabinitol-P2 was then added, and any exchange of [2'- C]carboxyarabinitol-P2 with the non-radioactive inhibitor was allowed to proceed for 5 hr. The contents of the vial were added to 2 mL of 30% polyethylene glycol containing 20 mM MgClz to precipitate the protein (Hall & Tolbert, 1979). The precipitate was collected by centrifugation, and washed with 3 mL 20% polyethylene glycol containing 20 mM MgC12, and recentrifuged. This washing procedure was repeated, the twice washed final precipitate was dissolved in 0.05 M Bicine buffer, pH 8.0, and an aliquot was taken for determination of radioactivity. An identically treated solution was used for the determination of enzyme activity (see Materials and Methods). A 15 s assay was performed in assay buffer containing no added Mg2+ and 10 mM NaH[14C]03. (A) The NaHCO3 concentration was varied in the preincubation step at a constant MgClz concentration (20 mM). The C02 concentration was calculated using pKa = 6.35 for the HC03, C02 equilibrium. Open data points represent non-exchangable [2'-14C]carboxyarabinitol-P that was precipitated with the enzyme, and the closed data points2 represent enzyme activity. The data were normalized by setting the maximal enzyme activity (Vma = 2.6 nmol/min-mg protein) and the maximal amount of bound [2 -f4C]carboxyarabinitol-P2 (n 7.2 mol ligand/mol protein) equal(to 1. {he solid line 15 drawn from Michaelis-Menten theory with V or n = 1 and K (002) = 14.5 uM. max max app (8) The MgCl concentration was varied in the preincubation step at a constant NaH 0 concentration of 10 mM. Open data points represent non-exchangeabIe [2'-14C]carboxyarabinitol-P2 that was precipitated with the enzyme, and the closed data points represent enzyme activity. The data were normalized by setting the values of the parameters at 10 mM MgCl2 (n = 7 mol ligand/mol protein; v = 2.5 umol/min-mg protein) equal to 1. 143 L 200 100 C02 (11M) ..o 5 0. 86666653354 6 693 28m LOO- Figure 5A 144 MgCl2 (mM) LOO" E nu 5 nu 33:04. 3 0525 6:283... Figure 58 145 the other carbons (data not shown). To examine the importance of the C-2' carboxyl group on the interaction of carboxyarabinitol-P2 with the enzyme, the hydroxymethyl and aldehyde analogues of the carboxypentitol bisphosphates were synthesized and their interactions with the enzyme were examined. Interaction of Ribulose-P; Carboxylase/oxygenase with Structural Analogues of the Carboxypentitol Bisphosphates. The hydroxymethyl derivatives of the carboxypentitol bisphosphates are simple competitive inhibitors with respect to ribulose-P2. The Ki values are 80 u" for hydroxymethylribitol-Pz and 5 uM for hydroxymethylarabini- tol-Pz. Neither compound exhibited a large time dependence in their inhibition of the enzyme. The enhanced binding of carboxyarabinitol-P2 (K < 10'11 M) as compared to that of hydroxymethylarabinitol-P2 (K1 = 5 x 10'6 M) might be due to an interaction of the negatively charged carboxyl group with either protein-bound Mgz+ or a positively charged amino acid group that is exposed when Mg2+ binds to the enzyme. A number of investigators have reported the presence of several lysine groups at the active site of ribulose-P2 carboxy- lase/oxygenase (Norton et al., 1975; Schloss & Hartman, 1977; Paech et al., 1977; Whitman & Tabita, 1978a,b; Paech & Tolbert, 1978). Since the branched-chain inhibitors mentioned are tightly bound to the enzyme, it appeared that aldehydo analogues of the carboxypentitol bisphOSphates might form covalent adducts in a very specific fashion with these active-site lysine groups. Although aldehydopentitol-Pz and 3,4-dideoxyaldehydopenti- tol-PZ inhibited the enzyme at low concentrations, attempts to detect 146 Schiff base formation between the enzyme and the 14C-labeled inhibitors by reduction with sodium borohydride or sodium cyanoborohy- dride at pH values from 6.5 to 8.5 in the presence or absence of C02 and Mg2+ were unsuccessful. The inhibition of the enzyme by these reagents was presumably due to binding via their negatively charged phosphate groups. Interaction of Ribulose-Pp Carboxylase[Oxygenase with Three Carbon Phosphate Esters. The inhibitors mentioned thus far have been designed to mimic the carboxylated intermediate of the reaction. As such they contain structural elements of both ribulose-P2 and C02. However, possible transition state analogues would also include those compounds whose structures more Closely resembles the product of the carboxylase reaction, D-glycerate-3-P. With this in mind, the effects of a number of these compounds were determined (Table 4). Neither dihydroxyacetone-P, O-glycerate-3-P, or L-glycerate-3-P were found to be highly inhibitory to ribulose-P2 carboxylase/oxy- genase. However, treatment of the enzyme with hydroxypyruvate-P resulted in severe inhibition of activity. It is interesting to note that the structural difference between hydroxypyruvate-P and dihydroxy- acetonE—P is the substitution of a carboxyl group for a hydroxymethyl group.' This difference causes a large change in the apparent binding constants for the two compounds, an effect that is reminiscent of the behavior of the branched chain acid- and hydroxymethyl-pentitolbisphos- phates. In addition, comparison of the relative affinities of the enzyme for D- or L-glycerate-3-P with that of hydroxypyruvate-P indi- cates the effect on enzyme binding of changing the hybridization (and geometry) about C-2 of these molecules. Possible reasons for the large 147 Table 4. Dissociation Constants for Three Carbon Sugar Phosphatesa Compound K1 (11M) D-Glycerate-B-P 840b L-Glycerate-3-P 900 Dihydroxyacetone-P «113,000 Hydroxypyruvate-P 20 aThe compounds listed were competitive inhibitors with respect to ribulose-P2. Inhibition constants (Ki) were determined by adding activated enzyme to a solution containing various concentrations of the sugar phosphate and ribulose-P . The concentrations of NaHC03 and MgCl used in the assay were 2 mM. Assays were conducted in duplicate at 30°C for 30 s. bFrom G.H. Lorimer, personal communication. differences in the binding constants for these compounds will be discussed shortly. Discussion From the studies presented here, it is clear that the “carboxy- ribitol-P2" used by earlier workers is a mixture of ribo and arabino isomers. The two isomers are similar to ribulose-P2 in their initial binding to the enzyme, but only the arabino isomer exhibits the second phase of tight binding. Therefore, the potent inhibition of ribulose-P2 carboxylase/oxygenase by "carboxyribitol-Pz" cited in earlier literature is due to carboxyarabinitol-P2.3 The slow, second phase of binding of carboxyarabinitol-P2 accounts for the 105-fold greater affinity of this ligand for the 148 enzyme relative to ribulose-P2. It is possible that carboxyarabini- tol-P2 mimics the transition state of the substrates and that the slow, tight binding of the analogue models a similar, but more rapid process that occurs with the transition state of the substrates. In the latter case, the rate of the binding process must equal or exceed the turnover rate of the enzyme. The difference between the rate of formation of the second binding state with carboxyarabinitol-P2 (0.04 5‘1) and the turnover rate of the enzyme active sites (b2.5 5‘1) may reflect the differences in structure between carboxyara- binitol-P2 and the transition state of the substrates or the carboxy- lated intermediate in the reaction. Exceedingly tight binding of an inhibitor to an enzyme has been used to infer a structural similarity to the transition state of the substrates in enzyme-catalyzed reactions (Wolfenden, 1972). However, the nature of the transition state in the carboxylation reaction is quite uncertain. Kinetic isotope experiments have been interpreted to imply that proton abstraction from C-3 of ribulose-P2 and proton addition to the (formal) carbanion of glycerate-3-P are slow relative to other catalytic events (Simon et al., 1964; Hurwitz et al., 1956; Fiedler et al., 1967). Additionally, nonlinear Arrhenius plots for the 3In the original experiments with carboxypentitol-P2 (Siegel & Lane, 1973), a purification procedure involving ion-exchange chroma- tography gave an asymmetric peak containing the carboxypentitol-P2 compounds. Only fractions around the peak's maximum were pooled and used for the inhibition studies. Our results indicate that these fractions probably contained an approximately 3:1 mixture of the two epimers, the larger fraction corresponding to carboxyarabinitol-P2. Preparation of the carboxypentitol-P2 compounds by other workers may have resulted in mixtures containing different proportions of the epimers. 149 carboxylation reaction may reflect the existence of different rate-- limiting steps at different temperatures (Bjorkman & Pearcy, 1971). Nevertheless, the 105-fold tighter binding of carboxyarabinitol-P2 relative to ribulose-P2 or carboxyribitol-Pz is consistent with the proposal that carboxyarabinitol-P2 more closely resembles the transi- tion state or the carboxylated intermediate of the reaction than does carboxyribitol-Pz. Accordingly, the intermediate in the enzyme-- catalyzed carboxylation of ribulose-P2 is most likely 2-C-carboxy-3-- keto-D-arabinitol-Pz (3) (Schemes 1 and 3). Studies with hydroxymethylarabinitol-P2 indicate that the carboxyl group of carboxyarabinitol-P2 is required for the second phase of inhibition. This second phase of binding which accounts for the 105-fold difference in binding between the hydroxymethyl and carboxy derivatives also requires C02 and MgZT, in agreement with reports that the enzyme active site, C02, Mg2+, and carboxypen- titol-P2 form a very slowly dissociating, quaternary complex (Miziorko, 1979). To the extent that this quaternary complex repre- sents the transition state or intermediate in the enzyme reaction, these results indicate a specific role for Mg2+ in the stabiliza- tion of the transition state of the substrates or of intermediate 3. In this regard, the recent findings (O'Leary et al., 1979) of a protein-bound carbamate resonance in the 13C NMR spectrum of Rhodospirillum rubrum enzyme incubated with KH[13C]O3 and the existence of a slowly exchanging species of C02 in the presence (Miziorko, 1979) and absence (Lorimer, 1979) of carboxypentitol-P2 support the model of enzyme activation (Lorimer et al., 1976) in which the negative charge of an enzyme-bound carbamate promotes the binding CHZoPog I HO- C- c: +Mg+ °t ‘0’ 150 H _ \. "N'I Av c=C (6'_M ys vs Ho/C g\ 12.3: HI H (5' i°é I... R I 2 HfiFOH CH20P03 ‘I3H2°P°a A C|=O f / HT H-CIS-OH R Scheme 3 151 of Mgz+ to form the active enzyme-carbamate-Mg2+ species (Chart 1, p 104). When the Mgz+ in this complex is replaced by Mn2+, a large effect on the longitudinal relaxation rate of a rapidly exchanging species of [13C]02 is noted, and a distance of 5.4 A between the Mn2+ and the rapidly exchanging C02 Species was calculated (Miziorko & Mildvan, 1974). Therefore, the close proximity of Mg2+ to both a slowly and rapidly exchanging species of C02, together with the requirement of Mg2+ and a negatively charged carboxyl group for the tight binding of carboxyarabinitol-P2 to the enzyme, indicates a role for Mg2+ in both activation of the enzyme and catalysis. Earlier reports indicating that Mg2+ is not required beyond the amount required for enzyme activation (Laing & Christeller, 1976) may be rationalized by concluding that the "activa- tor" Mg2+ and the "catalytic" Mg2+ are one and the same. The observation that Mn2+ interacts preferentially with the carboxyl group of carboxypentitol-P2 in solution indicates that a similar interaction may occur at the active site. Indeed, Miziorko (1978) has presented preliminary evidence that the environment of Mn2+ in the quaternary complex is distorted from the octahedral environment of Mn2+ in solution. This distortion may be explained by insertion of a new ligand into the inner coordination sphere of Mn2+. Whether the new ligand is from carboxypentitol-P2, enzyme, or both is unresolved, though the preferential binding of Mn2+ to the carboxyl group of carboxyarabinitol-P2 in solution supports the possibility that this ligand is one of the carboxyl oxygens. The hypothesis that the Mg2+ ion in the enzyme-carbamate-- Mg2+ complex stabilizes the carboxylated intermediate (3) provides 152 a possible explanation for the effect of Mn2+ in changing the par- titioning between the carboxylase and oxygenase reactions. Substitu- tion of Mn2+ for Mg2+ changes the carboxylase/oxygenase ratio (at 10 mM H003 or 0.25 mM 02) to 1.1 from 11 (Wildner & Henkel, 1978). If the enediol of ribulose-P2 is indeed the species that is attacked by both molecular oxygen and C02, then stabilization of intermediate §_by the metal should shift the carboxylase/oxygenase ratio in favor of the carboxylation reaction. The finding that Mn2+ shifts the ratio in favor of the oxygenation reaction may then be explained by asserting that intermediate §.is stabilized less well by an+ than by M92+. However reasonable these speculations may be, further experimentation is necessary to determine their validity. The conclusion that the intermediate in the carboxylation reaction has the arabino configuration has interesting mechanistic implications. If, in the cleavage of intermediate §_by H20, C-2 achieves a (formal) carbanion character, attack of hydroxide ion at C-3 and addition of a proton at C-2 would result in the formation of an equimolar mixture of D- and L-glycerate-B-P. In fact, solution studies revealed that the glycerate-3-P formed by nonenzymatic hydrolysis of g was a mixture of D- and L-glycerate-3-P and that the L-glycerate-3-P was derived from C-1, C-2 and C-2‘ of.§ (Siegel & Lane, 1973). In contrast, two mole- cules of D-glycerate-3-P are produced in the enzymatic reaction (Weissbach et al., 1956; Jakoby et al., 1956). This difference in the enzymatic and nonenzymatic reactions might be explained by postulating a glycerate-3-P epimerase activity in ribulose-P2 carboxylase/oxy- genase, a carbanion inversion mechanism, or other, less likely 153 mechanisms involving an electron-deficient carbon at C-2 of Q. We have shown that the enzyme does not epimerize L-glycerate-B-P (data not shown). Therefore, a mechanism involving carbanion inversion seems to be most probable. Since the formation and inversion of the C-2 carbanion of glycerate-B-P would be expected to be slow and thermodynamically unfavorable, a mechanism for its stabilization in the active site should exist. This stabilization may occur via the mechanism shown in Scheme 3. Addition of a proton to the front face of the double bond in fi_yields D-glycerate-B-P. According to this mechanism, the efficient conversion of.§ to two molecules of D-glycerate-3-P requires the stabi- lization of g. This stabilization might be conveniently accomplished by complexation of 3 with the Mg2+ required for enzymatic activity. The tight binding observed for hydroxypyruvate-P supports this hypo- thesis. Hydroxypyruvate-P is isosteric with 5, and therefore has the negative charges of the phosphate and carboxyl groups in the same relative positions as they are in 3, Therefore, hydroxypyruvate-P may also be viewed as a transition state analogue of g, The structural requirements for this three carbon transition state analogue are quite strict, since a change in geometry about C-2 or replacement of the carboxyl group with a hydroxmnethyl group greatly increases the apparent dissociation constants for these compounds. If Mgz+ does indeed form a stable complex with the aci-acid intermediate Q1), then these structural requirements are readily understandable. These data, therefore, suggest that the activation of the enzyme with C02 and Mg2+ provides a site for M92+ from which it can stabilize the intermediates.§ and g, Stereochemical control is then 154 exerted by the enzyme in the last step of the reaction, the addition of a proton to the aci-acid intermediate (1). CHAPTER TWO THE ROLE OF EFFECTOR MOLECULES IN THE ACTIVATION OF RIBULOSE BISPHOSPHATE CARBOXYLASE/OXYGENASE The activation of ribulose-P2 carboxylase/oxygenase by C02 and Mgz+ to form the active, ternary (ECM) form of the enzyme has been discussed previously (Literature Review). The knowledge that this form of the enzyme is the active form has greatly facilitated the kinetic analysis of the enzyme. However, it is not certain that the activation by these two species can account for the amount of enzyme activity that is observed under physiological conditions. This is because the binding constants of C02 and M92+ are such that under physiological conditions of pH, [002], and [MgZ+], the amount of enzyme in the active ECM form is expected to be only 18-20% of the total enzyme. Since this appears to be too low an activity to account for the observed rates of photosynthesis, a number of investigators have sought other effectors of enzyme activity. Various sugar phosphates have been shown to activate the enzyme at physiological concentrations of C02 and Mg2+ (See Literature Review). Activation by these compounds required preincubation of the effector with the enzyme. When the preincubation step was omitted, these compounds inhibited the enzyme. Some very complicated allosteric mechanisms have been preposed to account for these phenomena, and the physiological importance of the effectors has been widely discussed. 155 156 It may be argued, however, that the very high concentration of ribu- lose-P2 carboxylase/oxygenase in the chloroplast (3.2-4 mM in active sites) makes the idea of regulation of the enzyme by the usual meta- bolic models highly improbable. That is, since the effectors can never produce more active enzyme than that amount obtained under saturating concentrations of C02 and MgZ+, they are constrained to have a maximal physiological effect of increasing the enzyme activity 5-fold. This effect would require an effector concentration of greater than 3.2 mM. which is much greater than the physiological concentrations of the most active of the effectors. However, the possibility that a wide variety of effectors can exert an additive influence on the activation state of the enzyme has kept alive the argument of allosteric control. This chapter describes experiments designed to elucidate the mode of action of the various effectors and their possible physiological significance in the regulation of ribulose-P2 carboxylase/oxygenase. It is shown that the effectors exert their action by stabilizing the active ECM form of the enzyme through binding at the ribulose-P2 substrate site. This mode of action explains the observed requirement for preincubation of the effector with the enzyme, and casts doubt on the current models of physiological regulation by these compounds. Materials and Methods Sugar phosphates were synthesized (Part I, Chapter l) or obtained from Sigma Chemical Co. Enzyme preparations and general assay procedures were described in Chapter 1. 157 Enzyme Activation and Assay. Enzyme at 0.04-2 mg/mL was incubated in Bicine buffer (0.1 M NaT-Bicine, (pH 8.2), 0.1 mM EDTA, 1 mM dithiothreitol) that contained variable concentrations of MgClz, NaHC03, and effector. The incubation was allowed to proceed at 15°C or 30°C for at least 30 min prior to assay. When it was desired to measure only the activation of the enzyme by a particular effector, a 5-10 uL aliquot of the preincubated enzyme solution (1-2 mg/mL) was delivered into 0.49-0.495 mL Bicine solution of the appropriate compo- sition so that the final desired concentrations were achieved (NaH[14C]O3 = 1 mM; MgZ+ = 20 mM; ribulose-P2 = 0.5 mM). This procedure resulted in a 50-100 fold dilution of the effector so that any inhibitory effects of the effector in the assay were minimized. The assay was allowed to proceed for 15-30 s before 0.3 mL of 2N HCl was added to stop the reaction. All assays were performed in duplicate. Specific conditions are given in the text. Trapping of the Enzyme-COp-Mg2+ Complex with Carboxyarabini- 321232. Enzyme (1 mg/mL) was incubated at 30°C in Bicine buffer containing 1.0 mM NaH[14c303 (4.6 Ci/mol) and 20 mM MgCl2 for 30 min with or without a variable concentration of effector (total volume was 1 mL). A 10 uL aliquot was removed for the determination of enzymatic activity as described previously (0.5 mL total assay volume). Additionally, a 0.4 mL aliquot was delivered into 4.1 mL of a carboxyarabinitol-P2 solution (0.1 mM carboxyarabinitol-P2, 1 mM NaHC03, 20 mM MgClz in 50 mM Bicine buffer). The presence of a 10-fold excess of NaHC03 helped to dilute the specific radioactivity of the unbound NaH[14c103 ([14c102). After at least 30 min, 4.5 mL of a solution containing 40% polyethylene glycol, 5 mM 158 NaHC03, 20 mM MgCl2 in 0.1 M Bicine buffer was added to precipitate the enzyme-COZ—MgZT-carboxyarabinitol-P2 complex (Hall & Tolbert, 1979). The precipitate was collected by centrifugation (27,000 xg x 10 min) and washed with 4 mL of a 20% polyethylene glycol solution containing 5 mM NaHC03, 20 mM MgClz in 0.1 M Bicine buffer. The twice washed precipitate was then dissolved in 0.6 mL of 50 mM Bicine buffer containing 5 mM NaHC03, and an aliquot was taken for radioactivity determination by liquid scintillation counting. All assays and trapping experiments were performed in duplicate. When it was desired to vary the concentration of NaH[14C]03 in the activation step, the different amounts of NaH[14C]03 that were carried over into the assay solution were taken into account by enzymatically assaying the resulting specific activity of the assay NaH[14C]03 with a known amount of ribulose-P2. The enzymatic reaction was allowed to run to completion and the acid stable radioac- tivity observed divided by the amount of ribulose-P2 added was taken as the specific activity of the NaH[14C]03 ([14CJOZ) in the assay solution. A separate determination of the total radioactivity in the solution then allowed the calculation of the concentration of NaH[14C]03. Results Survey of Effectors. A wide variety of effectors have been shown to stimulate the activity of ribulose-P2 carboxylase/oxygenase upon preincubation of the effector with the enzyme (Buchanan & Schurmann, 1973a,b; Chu & Bassham, 1974,1975). This preincubation is required because activation is a slow process relative to the catalytic rate 159 of the enzyme. The rate limiting step for activation, in the absence of effector, has been shown (Lorimer et al., 1976) to involve the slow, reversible addition of C02 to a lysine e-amino group to form a carbamate. Complexation of Mg2+ with the EC form of the enzyme to form the active ECM form then proceeds rapidly. Since the breakdown of the ECM complex (in the presence of high concentrations of M92+ and ribulose-P2) is much slower than the time required to perform the assay for catalytic activity (15-60 5), the amount of the active enzyme may be measured by determining enzyme activity. The presence of effector does not change the rate of activation (data not shown). Therefore, it appears that the same process of carbamate formation may be involved in effector mediated activation, and the experimental protocol of preincubation followed by assaying for catalytic activity may be used for the determination of effector action. The list of possible effectors of enzyme activation has been extended to include a number of other compounds (Table 5). It can be seen from this listing that the structural requirements for an effector of enzyme activation are minimal. Molecules as different as inorganic salts, sugar monophosphates, and sugar diphosphates are effective activators. It appears that small monophosphate esters are effective activators whereas larger monophosphate esters either inhibit the activity slightly (ribose-S-P) or have little effect (glucose-G-P). Furthermore, for a given effector, the extent of activation may increase upon increasing the effector concentration (e.g., phosphate, glycerate-3-P, glycerol-Z-P), or it may decrease upon increasing the effector concentration (e.g., gluconate-G-P, carboxyarabinitol-P2, fructose-1,6-P2). This effect is actually an artifact of the assay 160 Table 5. Survey of Effectors of Ribulose-Pz Carboxylase/oxygenase Activationa Relative Enzyme Activity Effector Concentration Effector 1 mM 10 mM gluconate-6- -PC fructose-1, 6- -P2C NADPHc sodium phosphatec sodium sulfatec a—D- -glucose-1, 6- -P2 glycerate 3Tc phosphoenolpyruvate glycolate-Z-P glycerol-Z-P 2,3-diphosphoglycerate ATP 0 HHHHNQHANmHmwmm-kmmmm sedoheptulose-1,7T2c sodium nitrate glucose-G-P ribose-S- P 2-pentanone-1, 5— —P2C xylulose-P C carboxyribitol-P C hydroxypyruvate-Ec OOOOOOHOD—It-‘HHNHI—‘NNMHH ooooomtowawmooummmmwowh O O O HNmVOi-h AAA“ HNOOOOHHHHHHNO—‘Hl—‘HNNN O 3333 aEnzyme was activated at 1 mg/mL in 90 mM Bicine buffer, pH 8.2 with 1 mM NaHCO 20 mM MgCl2 and various effector concentrations. After 30 min at 30°C, a 20 uL aliquot was delivered into 0.48 mL of assay buffer so that the final concentrations of assay components were 71 ug/mL enzyme, 1 mM NaHEW‘4CJO 21 mM MgCl, 0. 7 mM ribulose-P2. Assays were conducted in triplicate for 3 5. Under these conditions, the specific activity of the enzyme in the absence of effector was 390 nmol/min-mg, and its activity when activated with 10 mM NaHC03 was éOOO nmol/min-mg. The relative activities are the ratio of the activity obtained in the presence of effector to that obtained in the absence of effector. cThese compounds are well known to inhibit the enzyme activity in a competitive manner with respect to ribulose-P2 when presented to the enzyme simultaneously with ribulose-P2. 161 and depends on the effector concentrations used and the binding con- stant of the effector. That is to say, the assumption used in deter- mining the stimulation of activity by an effector is that any increases in the amount of active enzyme may be measured by assaying catalytic activity with ribulose-P2. However, since a number of the effectors are also competitive inhibitors of the enzyme with respect to ribu- lose-P2, any carryover of these effectors into the assay solution may cause a decrease in enzyme activity due to the competitive inhibition and depending on the inhibition constant for the particular effector. Thus, an effector may indeed activate the enzyme but its effect may go unnoticed if sufficient effector is present in the assay solution to cause inhibition. This inhibition may be minimized by using high con- centrations of ribulose-P2 in the assay solution and very high dilu- tion when transferring the preincubated enzyme solution into the assay buffer. Thus, when gluconate-6-P is included at a concentration of 10 mM in the activation step and assayed with a 25 fold dilution, it appears to afford less of the activated enzyme than when its concentra- tion in the activation step is only 1 mM. However, under the same conditions, but with a ZOO-fold dilution into the assay medium, the 10 mM and 1 mM samples produce similar amounts of stimulation of enzyme activity (data not shown). The stimulatory effect of a given effector also appears to depend on its rate of dissociation from the enzyme. The discussion of this property will be deferred until after a descrip- tion of the mechanism of effector stimulation. Mechanism of Effector Action. A large body of data suggests that in the absence of effectors, enzyme activation proceeds by the slow, readily reversible formation of an enzyme-carbamate-Mg2+ (ECM) 162 species (See Chart 1 and Literature Review). It was possible that effectors exert their action by somehow increasing the equilibrium concentration of the ECM form at a given [C02] and [MgZ+]. A technique to determine the amount of ECM form of the enzyme in a given enzyme solution was therefore required to test this possibility. As shown in Chapter I, carboxyarabinitol-P2 forms a tight complex only with the active ECM form of the enzyme. In addition, Miziorko (1979) has shown that the ECM-carboxyarabinitol-Pz complex is inert to exchange of its C02 and Mg2+ components with solute species. It appeared then, that carboxyarabinitol-P2 could be used to assess the relative amount of the ECM form in a given enzyme solution. This expectation was supported by an experiment in which enzyme was incubated with varying concentrations of [14C]02 at constant Mg2+ concentration, and then assayed at constant C02 and M92+ concentrations. As expected, the amount of active enzyme increased with increasing preincubation concentrations of C02 (Figure 6). In addition, by trapping the ECM form of the enzyme with carboxyarabini- tol-P2 (see Materials and Methods), it was found that incorporation of a non-exchangeable, enzyme bound [14C]02 Species depended on the solution concentration of C02 in the same manner that enzyme activity depended on C02 concentration (Figure 6). This non- exchangeable [14C102 species represents the enzyme bound carbamate (Lorimer & Miziorko, 1980), and therefore the experiments of Figure 6 demonstrate the feasibility of using carboxyarabinitol-P2 to assay the amount of the ECM form of the enzyme in solution. When effectors were preincubated with enzyme, [14C]Oz, and MgZ+, enzyme activity was seen to depend on effector concentration 163 FIGURE 6. The effect of varying C02 concentrations on enzyme activa- tion and on the extent of trapping of an enzyme-C02 Species by carboxyarabinitol-P2. The MgClz concentration was 20 mM. Other details are given in the Materials and Methods section. The closed data points represent enzyme activity and the open data Roints repre- sent the amount of enzyme bound radioactivity from NaH[1 C]03 that was rendered nonexchangeable by treatment with carboxyarabin- itol-P2. The data were normalized by setting the maximal enzyme activity (Vmax = 1.0 umol/min-mg protein) and the maximal amount of enzyme bound radioactivity (nmax = 6 mol ligand/mol protein) equal to 1. The solid line is drawn from Michaelis-Menten theory with Vmax (or "max) = 1 and Kapp (HC03) = 1.6 mM. 164' 8. 6. 4. 2 O O O 0 80568353354 3 US venom IO (mM) Hcog 165 in the same manner that formation of the carbamate complex depended on effector concentration (Figures 7A and 78). It is interesting to note the different effects of carboxyribitol-Pz and xylulose-PZ. Both of these compounds are tight binding (Ki ca. 2 uM) competitive inhi- bitors of ribulose-P2 carboxylase/oxygenase (with respect to ribu- lose-P2), but carboxyribitol-Pz interacts with the enzyme via a rapid equilibrium process (Chapter I) whereas the binding and release of xylulose-Pz are slow processes (McCurry'& Tolbert, 1977). There- fore, preincubation of enzyme, 002, MgZT, and carboxyribitol-P2 results in the formation of an ECM-carboxyribitol-P2 species. Upon dilution into the assay buffer or into the carboxyarabinitol-P2 trapping solution, carboxyribitol-Pz dissociates from the enzyme (due in part to the dilution of the complex) leaving the ECM form. The activity assay and the trapping experiment then measure the amount of ECM form present, which is due to the ECM plus ECM-carboxyribitol-Pz forms that were present at the end of the preincubation period. This is because the binding of ribulose-P2 and carboxyarabinitol-P2 to the ECM form occurs more rapidly than the breakdown of the ECM complex (at high Mg2+ concentrations). In contrast, the slow dissociation of xylulose-Pz prevents its release during the time of the assay, and therefore xylulose-Pz appears to inhibit activation. In addition, its slow release from the enzyme prevents binding of carboxyarabini- tol-P2 and therefore xylulose-Pz appears to cause a decrease in the amount of the available ECM form as judged by trapping with carboxyara- binitol-Pz. Similar behavior is shown by another ribulose-P2 ana- logue, 2-pentanone-1,5-P2 (Table 5). Figure 7B shows the effects of two compounds that interact with the enzyme in a rapidly reversible 166 I 1 I I I’ll—l— r 1 I r IlA—F‘ A B . . 1: r? . H r; .3- 2.0 - ~ 2.0 , ‘ 0 <1: 8 v E? - '6 23 1 o . IO 1/ « '5 C .9 13 O I: H l l l l j I IO 20” so ‘ 4' a ” zo CRBP or XuBP (uM) Pi (mM) or 6—P-Gluconaie(melO) FIGURE 7. The effect of varying effector concentrations on enzyme activation and on the extent of trapping of an enzyme-C02 species by carboxyarabinitol-P2. The MgClz concentration was 20 mM. Other details are given in the text. All data are relative to a control containing no effector. The enzyme actixity in the cantrol was 0.38 umol/min-mg protein, and the amount of 1 C from NaH[ C303 that was rendered non-exchangeable by carboxyarabinitol-P2 in the control was 2.9 mol ligand/mol protein. The closed data points represent enzyme activity and the Open data points represent non-exchangable radioactivity. (A) Carboxyribitol-Pz (circles) and xylulose-Pz (triangles) concentrations were varied. (B) Gluconate— 6-P (circles) and inorganic phosphate (triangles) concentrations were varied. 167 fashion but with widely different affinities for the enzyme. Steady State Analyses of Effector Action. The previous experi- ments were performed by preincubating the enzyme for a long period of time prior to measuring the initial catalytic rate. Thus, the slow (hysteretic) response of catalytic activity to changes in C02 and Mg2+ concentrations was masked, and the initial rate of enzyme activity appeared to follow Michaelis-Menten kinetics. While this protocol is experimentally convenient, it may not occur*1_hvivg. To determine any possible physiological roles for effectors, it was desired to know not their effect on the initial catalytic rate of this hysteretic enzyme, but rather their effect on the steady state rate of carboxylation. Therefore, it was required to measure product formation over a long period of time, so that the slow bindinngrocess involved 10.225E activation and catalysis could reach their steady state levels. The experiments to be described were performed with C02, MgZ+, and ribulose-P2 concentrations that were chosen to approximate physiological concentrations. Inorganic phosphate was chosen as the effector since its high (3-4 mM) physiological concentration (Portis et al., 1977) makes it a likely candidate for physiological regulation of enzyme activity. The results of such an experiment using inorganic phosphate as the effector (Figure 8) show that while the presence of effector in the preincubation can certainly result in profound changes in initial catalytic rates, at long times the presence of effector has little effect on the rate of product formation. For instance, if enzmne is preincubated with C02, MgZT, and phosphate and the reaction is initiated with ribulose-P2 (line A in Figure 8), the initial 168 FIGURE 8. Time course of C0 fixation catalyzed by ribulose-P carboxylase/oxygenase after fiifferent preincubation regimes. A 0.7 mL solution of enzyme( (41 ug/mk) was preincubated in 0.1 M Bicine buffer (pH 8. 2) with 2.3 mM NaH [1 C]03 20 units of carbonic anhydrase, and varying concentrations of Mg+ , ribulose-P2, and/or inorganic phosphate at 30‘C in air tight vials. Additional preincubation components were: (A) 3.4 mM MgCl2, 5.7 mM inorganic phosphate; (8) 3.4 mM MgCl ; (C) no other adfiitions; (D) 0.54 mM ribulose-P2; 5. 7 mM inorganic phosphate; (E) 0.54 mM ribulose-P2. After 30 min, a 50 uL aliquot of a solution containing MgZT, inorganic phosphate, and/or ribulose-P2 was added to initiate the catalytic reaction. Aliquots (40 uL) were withdrawn and delivered into HCl at the indicated times to stop the reaction. The final concentra- tions of components in all assays were 2.1 mM NaH[1 4C103, 3.1 mM MgCl, 0. 5 mM ribulose-P 2’ and 5. 3 mM inorganic phosphate. The qualitative dotted line lS a control which was activated as in (B), but which was assayed in the absence of inorganic phosphate. The numbers in parentheses are the enzymatic rates that were observed during the 20-30 min period (in nmol C02 fixed/min-mg protein). 169 .E 2 o a. o _ / :1 \ B , / w— / / ON // So ./ / '5 E C o // /E ('9) l 20 time (min) 170 catalytic rate is 2.3 fold greater than when the enzyme is preincu- bated without phosphate and assayed under the same conditions (line B in Figure 8). However, the rate of product formation decreases with time and the catalytic rates between 20 and 30 min for both situations are within 10% of each other. (This decrease in catalytic rate is not due to excessive substrate depletion since the presence of carbonic anhydrase assured an adequate supply of C02, and the largest change in [ribulose-P2] (line A, Figure 8) was from 0.5-0.41 mM. Since Km (ribulose-P2) = 20 nM, this substrate loss could only account for a 0.8% loss in activity due to the lower [ribulose-P2] at the longer times. Furthermore, competitive inhibition by the product glycerate 3-P (K1 m 1 mM) is negligible under these conditions.) Preincubation with ribulose-P2 and C02 (lines 0 and E in Figure 8) affords lower catalytic rates. Again, this may be explained by noting that ribu- lose-P2 is very tightly bound to the catalytic site in the absence of Mg2+ (Wishnick et al., 1970) and its rate of release from the enzyme is very slow. (For instance, it is only partially released during a l hr gel filtration (Paech & Tolbert, 1978).) Therefore, the catalytic rate increases with time for those Situations in which ribu- lose-P2 was preincubated with the enzyme, and this increase in rate corresponds to the slow dissociation of ribulose-P2 from the enzyme and the corresponding activation of the enzyme by C02 and MgZ+. Presumably, at even longer times all of the final rates of the pro- cesses in Figure 8 would approach the same value. Regardless of whether this last presumption is true, it can be seen that the presence of inorganic phosphate has little effect on the catalytic rate between 20 and 30 min. 171 Discussion A simple model for effector mediated activation of ribulose-P2 carboxylase/oxygenase is shown in Scheme 4. slow slow ER :__—‘- 15: EC = ECM :2. ECMRC -—' Products lie—e 11...: .12. E = enzyme; C = carbamate or C0 ; M = M92+3 A a effector; R = ribquse-PZ Scheme 4 The model is based on the observations that the formation of the ECM complex occurs by the slow, ordered addition of C02 and Mg2+ (Lorimer et al., 1976), and that most (and maybe all) of the effectors are competitive inhibitors of the enzyme with respect to ribulose-P2. The model predicts the following characteristics of effector action which have been confirmed by experiment. 1. Effectors stimulate the activation of the enzyme only in the presence of C02 and Mgz+ (G.H. Lorimer, personal communication, and unpublished observations). The extent of stimulation by an effector should be highest at low C02 and Mg2+ concentrations, and nil at saturating C02 and MgZ+ concentrations. 2. Effectors tend to be competitive inhibitors with respect to ribulose-P2. 3. The rate of activation of the enzyme is independent of effector concentration. 172 4. Preincubation with effector should result in stimulation of enzyme activity, and simultaneous addition of effector and ribulose-P2 should result in inhibition of activity. 5. For effectors that bind to the enzyme in a rapidly reversible fashion, preincubation of enzyme, C02, and Mg2+ with effector may shift the equilibria towards those enzyme forms that contain Mgz+ and 002 (i.e., ECM and ECMA). Dilution into buffer containing ribulose-P2 then causes the dissociation of the ECMA form, and the amount of active enzyme is thereby increased over the case in which effector was absent in the preincubation. As a result, trapping of the ECM complex with carboxyarabinitol-P2 should demonstrate that carbamate formation is a function of effector concentraton. 6. The substrate, ribulose-P2, has the dual effects of a) binding tightly and slowly (relative to the catalytic rate) to the inactive form of the enzyme (Wishnick et al., 1970; Paech & Tolbert, 1978) and b) stabilizing the active ECM form of the enzyme (Laing & Christeller, 1976) by slowing the rate of dissociation of the enzyme-carbamate-- M92+ complex. Therefore, the presence of ribulose-P2 is expected to slow the rates of transitions between active and inactive enzyme forms. As a result, preincubation of the enzyme with ribulose-P2 in the absence of Mg2+ results in much lower proportions of active enzyme (see also Criddle & Dailey, 1980), and the approach to a constant steady state level of carboxylation takes longer than when ribulose-P2 is omitted in the preincubation step. In like manner, inhibition of activity is expected for effectors that dissociate very Slowly from the ECM complex. 173 7. Multiple effectors may show additive effects upon preincubation with the enzyme, depending on their concentrations and relative affinities for the enzyme. 8. Rapidly equilibrating effectors should have little or no effect on steady state levels of enzyme activity, providing that their concen- trations are not so high as to cause marked competitive inhibition. Summary. The experimental results and the considerations outlined above suggest that the only active form of the enzyme is the enzyme-- carbamate-Mg2+ complex. The role of the effector is therefore secondary and involves the stabilization of the active, ternary complex in much the same way that ribulose-P2 has been shown to stabilize this complex (Laing & Christeller,'1976). The stimulation of activity by the effectors is only apparent when the effectors are preincubated with the enzyme (because of the slow rate of carbamate formation) and when the effector concentration is diminished in the assay by high dilution (due to competitive inhibition by the effector). These dilu- tion and order of addition effects therefore maximize the apparent stimulation of enzyme activity by an effector. Although the quaternary ECMA complex contains the activated ECM complex, the ECMA form itself is catalytically inactive since the effector occupies the ribulose-P2 substrate site. Therefore, the steady state rate of carboxylation, which depends not only on the catalytic rate constants but also on the slow rate constants involved in formation of the ECM complex, should be essentially independent of the presence of effector (except in those cases where the effector concentration is high enough to cause severe 174 competitive inhibition). Of course, these compounds may well have other physiological properties that cause increases or decreases in steady state C02 fixation in plants. However, the results presented here give no suggestion that such changes in enzymatic activity could occur by the direct interaction of these effector molecules with ribulose-P2 carboxylase/oxygenase. CONCLUDING DISCUSSION The properties of ribulose-P2 carboxylase/oxygenase are of interest to a whole variety of researchers. Plant physiologists require a detailed understanding of its role in photosynthesis and photorespiration, and kineticists may find fruitful examples of complex enzyme behavior in its mechanism of action. The previous Chapters offer some modest proposals that may be useful in explaining some of the effects of C02, Mng, and various effector molecules on the enzyme's activity. However, critical and unanswered questions concern- ing this enzyme remain. For instance, the physiological concentrations of C02 and Mg2+ do not appear to be sufficient to allow the con- centration of active enzyme that is necessary to explain the observed rates of C02 fixation in whole plants. Based on the model of effec- tor action presented in Chapter 2, it is difficult to envision how these metabolite molecules could modulate the in yjyg_enzyme activity in a meaningful way. Therefore, the manner in which the enzyme remains appreciably activated Ba 1132 remains uncertain. For enzymologists, ribulose-P2 carboxylase/oxygenase presents a difficult problem in kinetic analysis. Although the slow rate of acti- vation has allowed experimenters to examine the enzyme by differentiat- ing between the activation and catalysis steps, a deeper understanding of the physiological response of the enzyme to various metabolites requires a more sophisticated treatment, which takes into account the hysteretic response of the enzyme to C02, MgZ+, and ribulose-P2. 175 176 Such Steady state analyses may be expected to appear in the near future. The oxygenase reaction of this enzyme has received much less attention than the carboxylase reaction. The reasons for this may be due in part to the difficulties inherent in the proper assay of its activity and in part to research prejudices regarding the seeming lack of "chemically interesting" cofactors normally associated with the oxygenases. However, the lack of pretty colors and interesting responses to microwave radiation appears not to detract from the enzmne's ability to cause the incorporation of a great deal of molecular oxygen into organic compounds. Luckily, the enzyme is an ideal candidate for physical and kinetic studies since it can be obtained in gram quantities of high purity. Furthermore, its anomalous behavior as an oxygenase is beginning to attract a number of researchers from diverse backgrounds. It appears then, that future studies may allow the determination of the mechanism of the oxygenase reaction. It is the hope of many that such a determination will result in the ability to manipulate the relative rates of carboxylation and oxygenation in lilQ: However, regardless of the outcome, the study of this most abundant enzyme will continue to be an active field of research. APPENDICES APPENDIX I ANALYSIS OF THE APPROXIMATE TREATMENT OF THE BLOCH EQUATIONS MODIFIED FOR CHEMICAL EXCHANGE Complete analyses of the phenomenological Bloch equations as modified for chemical exchange may be found in several references (Gutowsky & Holm, 1956; McConnell, 1958). In this Appendix the main interest will be in the approximations of the equations that become allowable in the special case when the two species undergoing exchange have very unequal equilibrium populations and which are separated by a large frequency difference in the NMR spectrum. In particular, we will be interested in the tautomeric exchange reaction of the furanose phosphates as an example of just such a Special case (see Table 4 and Figure 5A of Part I). The following discussion is an analysis of the errors that are associated with the approximate solutions to the NMR steady state equations, and the subsequent limitations that are imposed through the use of these approximations. The reader is referred to the analyses of Meiboom (1961) and Anet and Basus (1978) for a more complete description. For the reaction 177 178 where pa and pb are the fractional populations in sites A and B, respectively. The Gutowsky-Holm equation describing the effect of chemical exchange between A and B on the lineshape of their NMR spectrum is .m1=““+&Q%§+WJ . mm where v I ‘ 1((1/T2)2 - [lama +mb) «1112 + 13(wa-wb)2)+ l/TZ, o = new, +wb) - w - to, - pbxm, wbn. - DEG-”a +0013) 'w](1+ ZT/Tz) +;§(pa ‘ pb)(“’a -0013), I) I T "TaTb/(Ta TTb) . and v(w) is the absorption signal at the frequency m(in 5'1); ma and.mb are the respective resonance frequencies (in 5‘1); T2 is the transverse relaxation time of both species in the absence of exchange (assumed to be equal); and Ta = 1/ka and Tb = 1/kb. (A is a constant at a given field strength). This equation has two maxima at low exchange rates (at ma and ob) and one maximum at fast exchange rates (at w = Pawa + pbwb). For the special case in which the two Species undergoing exchange have very unequal populations (e.g. pb << pa) and which are separated by a large frequency difference, the linewidth of the main line will remain small compared to the frequency_difference between the two species, and will not move far from its original position. Under these conditions, terms in m higher than wz become increasingly negligible, and it can be shown (Anet & Basus, 1978) that the resulting linewidth of the major line is approximately Lorentzian over the full range of exchange rates with a linewidth at half height of 179 Aa = 2pb6(1/16 + 16)‘1 + A0 , (A3) where 6 =1na -¢nb, Aa is the observed linewidth, and A0 is the linewidth of the resonance in the absence of exchange. As the rate of exchange increases, this linewidth reaches its maximum value when a. -2p spa/(rs? +6 1 a: 0 = Jb (A4) at (1/16 +16 )2 which occurs at T = 1/6 or o. For T = 1/6, Agax = Pb‘S + Ao : (A5) which correSponds to the exchange rate ka + kb 3 I/T = 6 Since paka = pbkb, it follows that at the maximum linewidth ka = pba or upon converting to units of Hz, ka = 211(va - vb)pb . (A5) The rate constant in eqn A6 is that associated with the maximum possible linewidth of the A resonance. For much lower rates of exchange, (when 16 >> 1) further approximations are available. In these cases, r5 >> 1/15, and eqn A3 becomes D II a 2pb/q‘q'Ao 2pb10%) until the rate of exchange causes an increase in linewidth to approximately 50% of the maximum linewidth expected. Eqn A6 suggests that this limitation will be most severe for those compounds that have an exceedingly small fraction of the free 181 carbonyl form (i.e., ribose-S-P and arabinose-S-P). However, by remaining within the imposed limits, the approximate solution may be used to derive the ring opening rates for the tautomerization reaction. Table A1: Comparison of Exact and Approximate Lineshape Analysesa a-Fructose-1,6T2b T kao pb Au An Ad % error 1°C) (s-l) 323-) (eqn AS) (eqn 42) (eqn A7) 20 4.2 0.086 140 2.332 2.337 0.2 38 27 0.12 195 9.597 9.594 0.03 8-Fructose-1,6-P2b T kBo pb Agax A3 A8 % error (‘9) (5‘1) 333' (eqn A5) (eqn A2) (eqn A7) 20 19 0.014 24.4 7.003 7.048 0.6 38 90 0.023 39.5 25.42 29.55 12 aThe chemical shifts are 8 (101 ppm), a (105 ppm) and ketone (212 ppm). The entries are calculated for Spectra obtained at 15.08 MHz with 1/"Tz = A bA Values are = 1 Hz. APPENDIX II PUBLICATIONS AND COPYRIGHT ACKNOWLEDGEMENTS As mentioned in the acknowledgements, a number of the studies presented in this dissertation were the result of collaboration with other workers. Some of the collaborative efforts resulted in copyrighted journal publications which include a number of previous Tables and Figures. References to these materials are given below. Publications John Pierce, N.E. Tolbert, and Robert Barker. Interaction of ribulose biSphosphate carboxylase/oxygenase with transition state analogues. Biochemistry 19, 934-942 (1980). John Pierce, N.E. Tolbert, and Robert Barker. A mass spectrometric analysis of the reactions of ribulose bisphosphate carboxylase/oxygenase. J. Biol. Chem. 255, 509-511 (1980). Anthony S. Serianni, John Pierce, and Robert Barker. Carbon-13-enriched carbohydrates: Preparation of triose, tetrose, and pentose phosphates. Biochemistry 18, 1192-1199 (1979). Stephen D. McCurry, Nigel P. Hall, John Pierce, Christian Paech, and N.E. Tolbert. Ribulose-l,S-bisphosphate carboxylase/oxygenase from Parsley. Biochem. Biophys. Res. Commun. 84, 895-900 (1978). Christian Paech, John Pierce, Stephen D. McCurry, and N.E. Tolbert. Inhibition of ribulose bisphosphate carboxylase/oxygenase by ribulose bisphosphate epimerization and degradation products. Biochem. Biophys. Res. Commun. 83, 1084-1092 (1978). Christian Paech, Stephen D. McCurry, John Pierce, and N.E. Tolbert. Active site of ribulose-1,S-bisphosphate carboxylase/oxygenase. in Photosynthetic Carbon Assimilation, ed. H.W. Siegelman and G. Hind, Plenum 227-243 (1978). 182 183 Abstracts A.S. Serianni, H.A. Nunez, J. Pierce, P.R. Rosevear, E.L. Clark, and R. Barker. Isotopically Enriched Carbohydrates: Synthesis and use in chemical and biochemical studies. Abstract RAMP-10, Am. Soc. Mass Spg§6)28th Amer. Conference on Mass Spec. and Allied Topics 1 . J. Pierce, N.E. Tolbert, and R. Barker. Interaction 0f ribulose-P2 carboxylase/oxygenase with transition state analogues. Abstracts of the 178th Am. Chem. Soc. Meeting, Biol. 2 (1979). J. Pierce, C. Paech, S.D. McCurry, and N.E. Tolbert. Inhibition of RuBP carboxylase/oxygenase by RuBP epimerization and degradation products. Abstract 14. Brookhaven Symp. on Photosynthetic Carbon Assimilation (1978). BIBLIOGRAPHY BIBLIOGRAPHY Acree, T.E. (1968) Doctoral Thesis, Cornell University. Alderfer, J.L. & Ts'o, P.0.P. (1977) Biochemistry 16, 2410. Anet, F.A.L. & Basus, V.J. (1978) J. Magn. Reson. 32, 339. Angyal, S.J. (1969) Angew. Chem. 8, 157. Angyal, S.J. & Pickles, V.A. (1972) Austral. J. Chem. 25, 1695. Atherton, F.R., Openshaw, H.T. & Todd, A.R. (1945) J. Chem. Soc., 385. Austin, P.W., Hardy, F.E., Buchanan, J.G. & Baddiley, J. (1963) J. Chem. Soc., 5350. Axelrod, B. & Jang, R. (1954) J. Biol. Chem. 299, 847. Badger, M.R. & Andrews, T.J. (1974) Biochem. Biophys. Res. Commun. 60, 204. Badger, M.R. & Lorimer, G.H. (1976) Arch. Biochem. Biophys. 115, 723. Baer, E. & Stanier, H.C. (1953) J. Am. Chem. Soc. 15, 4510. Bahr, J.T. & Jensen, R.G. (1974a) Plant Physiol. 53, 39. Bahr, J.T. & Jensen, R.G. (1974b) Arch. Biochem. Biophys. 164, 408. Bailey, J.M., Fishman, P.H. & Pentchev, P.G. (1970) Biochemistry g, 1189. Bailly, M.C. (1942) Bull. Soc. Chim. (France) 9, 314, 340, 405, 421. Bailly, 0. (1916) Ann. Chim. (Paris) 6, 133. Ballou, C.E. (1963) Methods Enzymol. 6, 479. Ballou, C.E. & Fischer, H.0.L. (1955) J. Amer. Chem. Soc. 11, 3329. Ballou, C.E., Fischer, H.0.L. & MacDonald, D.L. (1955) J. Amer. Chem. Soc. 71, 5967. Barker, S.A., Bourne, E.J., Pinkard, R.M. & Whiffen, D.H. (1958) Chem. Ind. (London), 658. Bartlett, G.R. (1959) J. Biol. Chem. 234, 459. 184 185 Bell, R.P. (1966) Adv. Phys. Org. Chem. 1, 1. Benkovic, P.A., Bullard, W.P., DeMaine, M., Fishbein, R., Schray, K.J., Steffens, J.J. & Benkovic, S.J. (1974) J. Biol. Chem. 888, 6047. Benkovic, S.J., Engle, J.L. & Mildvan, A.S. (1972) Biochem. Biophys. Res. Commun. 81, 852. Bensadoun, A. & Weinstein, D. (1976) Anal. Biochem. 18, 241. Bhagwat, A.S. & Sane, P.V. (1978) Biochem. Biophys. Res. Commun. 88, 865. Binsch, G. (1968) T0p in Stereochem. 8, 1968. Binsch, G. (1975) in "Dynamic NMR Spectroscopy", ed. L.M. Jackman & F.A. Cotton, Academic Press, 45. Bjorkman, 0.E. & Pearcy, R.W. (1971) Carnegie Inst. Washington, Yearb. 70, 511. Blackmore, P.F., Williams, J.F. & Macleod, J.K. (1976) FEBS Letters 81, 222. Branden, R. (1978) Biochem. Biophys. Res. Commun. 81, 539. Branden, R., Nilsson, T. & Styring, S. (1980a) Biochem. Biophys. Res. Commun. 88, 1297. Branden, R., Nilsson, T., Styring, S. & Angstrom, J. (1980b) Biochem. Bi0phys. Res. Commun. 92,1306. Brown, 0., Hayes, F. & Todd, A. (1957) Ber. 88, 936. Buchanan, B.B. & Schurmann, P. (1973a) J. Biol. Chem. 888, 4956. Buchanan, B.B. & Schurmann, P. (1973b) Curr. Top. Cell. Regul. Z, 1. Bunton, C.A., Llewellyn, D.R., Oldham, K.G. & Vernon, C.A. (1958) J. Chem. Soc., 3574. Bunton, C.A. & Shiner, V.J., Jr. (1961) J. Am. Chem. Soc. 88, 42. Byrne, W.L. & Lardy, H.A. (1954) Biochem. Bi0phys. Acta 18, 495. Calvin, M. (1954) Fed. Proc., Fed. Am. Soc. Exp. Biol. L8, 697. Campbell, 1.0., Dobson, C.M., Ratcliffe, R.G. & Williams, R.J.P. (1978) J. Magn. Reson. 88, 397. Capon, B. & Walker, R.B. (1974) J.C.S. Perkin II, 1600. Chollet, R. & Anderson, L.L. (1976) Arch. Biochem. Bi0phys. 118, 344. 186 Chollet, R. & Anderson, L.L. (1977) Biochim. Biophys. Acta 588, 228. Chollet, R., Anderson, L.L. & Housepian, L.L. (1975) Biochem. Biophys. Res. Commun. 88, 97. Chu, D.K. & Bassham, J.A. (1974) Plant Physiol.‘§£, 556. Chu, D.K. & Bassham, J.A. (1975) Plant Physiol. 88, 720. C00per, T.G., Filmer, 0., Wishnick, M. & Lane, M.D. (1969) J. Biol. Chem. 888, 1081. Dawson, R.M.C., Elliot, D.C., Elliot, W.H. & Jones, K.M. (1968) "Data for Biochemical Research, 2nd edition, Clarendon Press, Oxford, p. 103. Dixon, J.S. & Lipkin, D. (1954) Anal. Chem. 88, 1092. Eisenberg, 0., Baker, T.S., Suh, S.W. 8 Smith, W.W. (1978) in "Photosynthetic Carbon Assimilation", ed. Siegelman & Hind, Plenum Press, New York, p. 271. Ernst, R.E. (1966) Adv. Magn. Reson. 8, 1. Ferrier, R.J. (1962) J. Chem. Soc., 3544. Fiedler, F., Mullhofer, G., Trebst, A. & Rose, I.A. (1967) Eur. J. Biochem. 1, 395. Fishbein, R., Benkovic, P.A., Schray, K.J., Siewers, I.J., Steffens, J.J. & Benkovic, S.J. (1974) J. Biol. Chem. 888, 930. Forsen, S. & Hoffman, R.A. (1963) J. Chem. Phys. 88, 2892. Forsen, S. & Hoffman, R.A. (1964) J. Chem. Phys. 89, 1189. Frey, W.A., Fishbein, R., de Maine, M.M. & Benkovic, S.J. (1977) Biochemistry 88, 2479. Gallopo, A.R. & Cleland, W.W. (1979) Arch. Biochem. Biophys. 188, 152. Gilham, P.T. & Tener, G.M. (1959) Chem. and Ind. (London), 542. Gorin, P.A.J. & Mazurek, M. (1976) Carbohydr. Res. 58, 171. Gray, G.R. (1971) Biochemistry 18, 4705. Gray, G.R. and Barker, R. (1970) Biochemistry 8, 2454. Grimshaw, C.E., Whistler, R.L. & Cleland, W.W. (1979) J. Amer. Chem. Soc. 181, 1521. 187 Grindley, T.B., Gulasekharam, V. & Tulshian, 0.8. (1977) Abstract of Pa ers, 2nd Joint Conference of the Chemical Institute of Canada ang the American Chemical Society, Division of Carbohydrate Chemistry, No. 12, Montreal, May 29-June 2. Gutowsky, H.S. & Holm, C.H. (1956) J. Chem. Phys. 88, 1228. Gutowsky, H.S. & Saika, A. (1953) J. Chem. Phys. 88, 1688. Hall, N.P. & Tolbert, N.E. (1979) FEBS Letters 88, 167. Hamilton, G.A. (1974) in "Molecular Mechanisms of Oxygen Activation", ed. 0. Hayaishi, Academic Press, New York, p. 405. Hartman, F.C. and Barker, R. (1965) Biochemistry g, 1068. Hartman, F.C., Norton, I.L., Stringer, C.D. & Schloss, J.V. (1978) in "Photosynthetic Carbon Assimilation" ed. Siegelman & Hind, Plenum Press, New York, p. 245. Harvey, W.E., Michalski, J.J. & Todd, A.R. (1949) J. Chem. Soc. 815. Hodge, J.E. & Hofreiter, B.T. (1962) Methods Carbohydr. Chem. 8, 380. Hoffman, R.A. & Forsen, S. (1966) J. Chem. Phys. 88, 2049. Horecker, B.L., Hurwitz, J. & Weissbach, A. (1958) Biochem. Prep. 8, 83. Hurwitz, J., Jakoby, W.B. & Horecker, B.L. (1956) Biochim. Biophys. Acta 88, 194. Isbell, H.S. & Pigman, W. (1969) Adv. Carbohydr. Chem. 81, 14. Jakoby, W.B., Brummond, 0.0. & Ochoa, S. (1956) J. Biol. Chem. 888, 811. Jensen, R.G. & Bahr, J.T. (1977) Ann. Rev. Plant Physiol. 88, 379. Jensen, R.G. & Bassham, J.A. (1966) Proc. Nat. Acad. Sci. USA 88, 1095. Kawashima, N. & Wildman, S.G. (1970) Ann. Rev. Plant Physiol. 88, 325. Kiliani, H. (1887) Ber. 88, 339. Klybas, V., Schramm, M. and Racker, E. (1959) Arch. Biochem. Bi0phys. 80, 229. Laing, W.A. & Christeller, J.T. (1976) Biochem. J. 888, 563. Laing, W.A., Ogren, W.L. & Hageman, R.H. (1974) Plant Physiol. 88, 678. 188 Lampson, G.P. & Lardy, H.A. (1949) J. Biol. Chem. 888, 693. Lapper, R.D. & Smith, I.C.P. (1973) J. Am. Chem. Soc. 88, 2880. Lapper, R.D., Mantsch, H.H. & Smith, I.C.P. (1973) J. Amer. Chem. Soc. 88, 2878. Lawlis, V.B. & McFadden, B.A. (1978) Biochem. Biophys. Res. Commun. 88, 580. Lee, J.B. (1963) J. Org. Chem. 88, 2473. Leloir, L.F. & Cardini, C.E. (1957) Methods Enzymol. 8, 840. Lenz, R.W. & Heeschen, J.P. (1961) J. Polymer Sci. 88, 247. Levene, P.A. & Steller, E.T. (1933) J. Biol. Chem. 888, 187. Lewis, K.F. & Weinhouse, S. (1957) Methods Enzymol. 8, 269. Lorimer, G.H. (1978) Eur. J. Biochem. 88, 43. Lorimer, G.H. (1979) J. Biol. Chem. 885, 5599. Lorimer, G.H. & Andrews, T.J. (1973) Nature 888, 359. Lorimer, G.H., Andrews, T.J. & Tolbert, N.E. (1973) Biochemistry 88, 18. Lorimer, G.H., Badger, M.R. & Andrews, T.J. (1976) Biochemistry 88, 529. Lorimer, G.H., Badger, M.R. & Heldt, H.W. (1978) in "Photosynthetic Carbon Assimilation,“ ed. Siegelman & Hind, Plenum Press, New York, p. 283. ' Lorimer, G.H. & Miziorko, H.M. (1980) Biochemistry, in press. Loring, H.S., Levy, L.W., Moss, L.K. & Ploeser, J.M. (1956) J. Amer. Chem. Soc. 88, 3724. Los, J.M., Simpson, L.B. & Wiesner, K. (1956) J. Amer. Chem. Soc. 88, 1564. Lowry, T.M. & John, W.T. (1910) J. Chem. Soc. 88, 2634. Maehr, H. & Smallheer, J. (1978) Carbohydr. Res. 88, 178. Mann, B.E. (1976) J. Magn. Reson. 88, 17. Mann, B.E. (1977) Prog. in NMR Spectr. _1_;_, 95. McConnelI, H.M. (1958) J. Chem. Phys. 88, 430. 189 McCurry, S.D. (1979) Doctoral Thesis, Michigan State University. McCurry, S.D. & Tolbert, N.E. (1977) J. Biol. Chem. 888, 8344. McCurry, 5.0., Hall, N.P., Pierce, J., Paech, C. & Tolbert, N.E. (1978) Biochem. Biophys. Res. Commun. 81, 895. McDaniel, D.H. & Smoot, C.R. (1956) J. Phys. Chem. 88, 966. Meyerhof, O. & Lohmann, K. (1934) Biochem. Z 818, 89. Michelson, A.M. & Todd, A.R. (1949) J. Chem. Soc., 1949. Midelfort, C.F., Gupta, R.K. and Rose, I.A. (1976) Biochemistry 88, 2178. Miziorko, H.M. (1978) in "Photosynthetic Carbon Assimilation" ed. Siegelman & Hind, Plenum Press, New York, p. 41. Miziorko, H.M. (1979) J. Biol. Chem. 881, 270. Miziorko, H.M. & Mildvan, A.S. (1974) J. Biol. Chem. 818, 2743. Moffatt, J.G. and Khorana, H.G. (1957) J. Amer. Chem. Soc. 18, 1194. Mullhofer, G. & Rose, I.A. (1965) J. Biol. Chem. 838, 1341. Nef, J.U. (1910) Ann. 888, 1. Norton, I.L., Welch, M.H. & Hartman, F.C. (1975) J. Biol. Chem. 250, 8062. O'Leary, M.H., Jaworski, R.J. & Hartman, F.C. (1979) Proc. Natl. Acad. Sci. U.S.A. 88, 673. Paech, C. & Tolbert, N.E. (1978) J. Biol. Chem. 888, 7864. Paech, C., McCurry, S. 0., Pierce, J. & Tolbert, N. E. (1978) in "Photosynthetic Carbon Assimilation" ed. Siegelman & Hind, Plenum Press, New York, p. 227. Paech, C., Ryan, F.J. & Tolbert, N.E. (1977) Arch. Biochem. Bi0phys. 179, 279. Paulsen, J.M. & Lane, M.D. (1966) Biochemistry 8, 2350. Perlin, A.S., Cyr, N., Ritchie, G.S. & Parfondry, A. (1974) Carbohydr. Res. 88, C1. Petersson, G. (1970) Tetrahedron 88, 3413. Pierce, J., Tolbert, N.E. & Barker, R. (1980) J. Biol. Chem. 888, 509. Pigman, W. & Isbell, H.S. (1968) Adv. Carbohydr. Chem. 88, 11. 190 Pon, N.G., Rabin, B.R. & Calvin, M. (1963) Biochem. Z 888, 7. Portis, A.R., Chon, C.J., Mosbach, A. & Heldt, H.W. (1977) Biochim. Biophys. Acta 888, 313. Que, L. & Gray, G.R. (1974) Biochemistry 88, 146. Rabin, B.R. & Trown, P.W. (1954) Nature 808, 1290. Reynolds, S.J., Yates, D.W. & Pogson, C.I. (1971) Biochem. J. 888, 285. Ritchie, R.G.S., Cyr, N., Korsch, B., Koch, H.J., and Perlin, A.S. (1975) Can. J. Chem. 88, 1424. Robison, P.D. & Tabita, F-.R. (1979) Biochem. Biophys. Res. Conmun. 88, 85. Rutner, A.C. & Lane, M.D. (1967) Biochem. Biophys. Res. Commun. 88, 531. Ryan, F.J. & Tolbert, N.E. (1975) J. Biol. Chem. 888, 4229. Schaffer, R. & Isbell, H.I. (1963) Meth. Carbohydr. Chem. 8, 11. Schloss, J.V. & Hartman, F.C. (1977a) Biochem. Biophys. Res. Commun. 75, 320. Schloss, J.V. & Hartman, F.C. (1977b) Biochem. Biophys. Res. Commun. 77, 230. Schloss, J.V., Norton, I.L., Stringer, C.0. & Hartman, F.C. (1978a) Fed. Proc. 88, 1300. Schloss, J.V., Stringer, C.D. & Hartman, F.C. (1978b) J. Biol. Chem. 888, 5707. Schowen, R.L. (1972) Prog. Phys. Org. Chem. 8, 275. Schray, K.J. and Benkovic, S.J. (1978) Acc. Chem. Res. 8, 136. Schwarz, J.A. & Perlin, A.S. (1972) Can. J. Chem. 88, 3667. Scott, A.I. (1964) "Interpretation of the Ultraviolet Spectra of Natural Products," Pergamon Press, Oxford, p. 36. Serianni, A.S. (1980) Doctoral Thesis, Michigan State University. Serianni, A.S., Nunez, H.A. & Barker, R. (1979a) Carbohydr. Res. 18, 71. Serianni, A.S., Clark, E. & Barker, R. (1979b) Carbohydr. Res. 88, 79. 191 Siegel, M.I. & Lane, M.D. (1972) Biochem. Biophys. Res. Commun. 58, 508. Siegel, M.I. & Lane, M.D. (1973) J. Biol. Chem. 818, 5486. Siegel, M.I., Wishnick, M. & Lane, M.D. (1972) The Enzymes 8, 169. Siegelman, H.W. & Hind, G. (1978) Eds. "Photosynthetic Carbon Assimilation", Plenum Press, New York. Simon, H., Dorrer, H.D. & Trebst, A. (1964) Z. Naturforsch B 88, 734. Sjodin, B. & Vestermark, A. (1973) Biochim. Bi0phys. Acta 881, 165. Smith, M., Moffatt, J.G. 8 Khorana, H.G. (1958) J. Am. Chem. Soc. 88, 6204. Somogyi, M. (1945) J. Biol. Chem. 888, 69. Spellman, M., Tolbert, N.E. & Hartman, F.C. (1979) Abstracts of Papers, 178th National Meeting of the American Chemical Society, Washington, D.C., September, 1979, BIOL. 3. Stringer, C.D. & Hartman, F.C. (1978) Biochem. Biophys. Res. Commun. 88, 1043. Sue, J.M. & Knowles, J.R. (1978) Biochemistry 81, 4041. Sugiyama, T., Nakayama, N., Ogawa, M. & Akazawa, T. (1968) Arch. Biochem. Biophys. 888, 98. Trontham, D.R., McMurray, G.H. & Pogson, C.E. (1969) Biochem. J. 881, 19. Trown, P.W. (1965) Biochemistry 1, 908. Trown, P.W. & Rabin, B.R. (1964) Proc. Nat. Acad. Sci. U.S.A. 88, 88. Weissbach, A., Horecker, B.L. & Hurwitz, J. (1956) J. Biol. Chem. 888, 795. Whitman, W.B. & Tabita, F.R. (1978a) Biochemistry 81, 1282. Whitman, W.B. & Tabita, F.R. (1978b) Biochemistry 11, 1288. Wildner, G.F. (1976) Ber. Dtsch. Bot. Ges. 88, 349. Wildner, G.F. & Henkel, J. (1977) Z. Naturforsch. C 88, 226. Wildner, G.F. & Henkel, J. (1978) FEBS Letters 88, 99. Wilkinson, G.N. (1961) Biochem. J. 88, 324. 192 Wishnick, M., Lane, M.D. & Scrutton, M.C. (1970) J. Biol. Chem. 888, 4939. Wishnick, M., Lane, M.D., Scrutton, M.C. & Mildvan, A.S. (1969) J. Biol. Chem. 244, 5761. Wolfenden, R. (1972) Acc. Chem. Res. 8, 10. Wurster, B. & Hess, B. (1974) FEBS Letters 88, 257.