Nuclear Magnetic Resonance Spectroscopy of Chlorophylls and Corrins Joseph J

REVIEWS

Nuclear Magnetic Resonance Spectroscopy of Chlorophylls and Corrins Joseph J. Katz'and Charles E. Brown* *Chemistry Division Argonne National Laboratory Argonne, III. 60439 USA * Department of Biochemistry The Medical College of Wisconsin Milwaukee, Wis. 53226 USA

Page i ntroduction 3

The ChlorophylIs k A. Structural Features k B. Experimental Aspects 5 C. Chemical Shift Assignments 5 D. Applicat ions of NMR 20

11 Corr ins A. Structural Features 32 B. *H NMR Chemical Shift Assignments and Applications 32 C. 13C NMR Chemical Shift Assignments and Applications 33 D. NMR Studies with other Nuclei U2 E. Summary

References

i. INTRODUCTION systems involved in important isomerization and methyl group transfer reac- The role of NMR in the study of com- tions, and are intimately involved in pounds of biological importance is protein and probably also in lipid and widely recognized and appreciated. carbohydrate metabolism. Vitamin B12 There can be few categories of such plays a very important part in hemo- substances, however, to which NMR has poiesis (stimulation of red blood cell made such substantive contributions as formation), and together with folic it has to the understanding of the acid participates in the formation of chlorophylls and the corrins. The deoxyribonucleotides from ribonucleo- chlorophylls are indispensible agents tides. Applied to the chlorophylls, in the conversion of the energy of NMR has provided information relevant light to chemical oxidizing and reduc- to such aspects as the sites of ing capacity. The natural corrins are exchangeable hydrogen, keto-enol tau- coenzymes for a number of enzyme tomerism, the biosynthetic pathways of

Vol. 5, No. 1/2 chlorophyll formation, hyperfine interactions in chlorophyll cations and other paramagnetic chlorophyll species, as well as more conventional NMR information useful in establishing the chemical identities of several chlorophylls of previously unknown structure. Per- haps the most important contribution from NMR to an understanding of chlorophyll behavior is the delineation of the coordination donor-acceptor properties of the chlorophylls that largely determine the state of chlorophyll j_n vivo. For the corrins, clarification of the path of biosynthesis and the variables that affect their coenzyme activities has up to now been the most significant contribution from NMR studies.

It. JHE CHLOROPHYLLS

The chlorophylls constitute a small group of closely related compounds whose function in nature is to collect light quanta and to use the subsequent electronic excitation energy to effect charge separation. The oxidizing and reducing power so produced is then used Figure 1. Structure and numbering of to drive redox reactions that would the chlorophylls. J_, Chlorophyll a (Chi otherwise not proceed spontaneously. a); 2, pyrochlorophyl1 a (Pyrochl a); Useful introductions to the role of 2, chlorophyll b (Chi b); k, bacter- chlorophyll in photosynthesis have been iochlorophyl1 a (Bchl a). The methyl provided by Govindjee and Rabinow itch chlorophy11 ides have the same macrocy- (1), by Clayton (2), and by Govindjee cle as the chlorophylls, but the phytyl (3). moiety is replaced by -CH3. The removal of the central Mg and its replacement A. Structural Features by 2H from chlorophylls and methyl chlorophyl1 ides forms pheophytins The chlorophylls are cyclic tetra- (Pheo) and methyl pheophorbides, pyrroles, and thus belong to the por- respectively. Chlorophylls £1 and £2 phyrin family. There are both important have an acrylic acid side chain at similarities and differences between position 7 in structure 1; in Chi £2, a the chlorophylls and the more widely vinyl group is also present at position studied porphyrins. The side chains of k. Both Chi £1 and £2 lack any esterif- the tetrapyrrole macrocycle , i.e. ying alcohol at position JZ. Bacter- methyl, ethyl, vinyl, and propionic iochlorophylI b has only an ethylidene acid, are much the same in both porphy- group, =CH-CH3 at position k in struc- rins and chlorophylls. The side chain ture J». Protochlorophyl I a is identi- positions likewise for the most part cal with J^ except for the absence of appear to be identical, arguing for protons 7 and 8. The protons at posi- similar biosynthetic pathways. The tions 7 and 8 are also missing in Chi chlorophylls, however, all have an ali- ci and C2. cyclic 5~membered ring V (Figure 1), which contains a keto carbonyl function at position C-9. Most of the chlorophylls contain a carbomethoxy group at

Bulletin of Magnetic Resonance the 10 position, but in several The susceptibility of chlorophylls important chlorophylls the carbomethoxy to oxidation by molecular oxygen neces- group is replaced by an H atom. Ring V sitates special precautions in record- is a structural feature unique to the ing NMR spectra. Reaction with oxygen chlorophylls. It is this feature that in polar organic solvents, particularly is mainly responsible for the rich and methanol , rapidly produces sufficient complex chemistry characteristic of the allomerized chlorophylls to complicate chlorophylls. The proton at C-10 in spectral interpretation. The allomer- chlorophylls containing a carbomethoxy ized chlorphylls are similar in chemi- group is part of a 3~keto ester system, cal structure to chlorophyll itself, which can undergo enolization. Enoliza- and the spectrum of a mixture of tion is associated with epimerization closely related but not identical com- at the chiral center at C-10 , and is pounds may show broad, poorly resolved implicated in a very complicated set of resonance peaks. Even 1% by weight of a oxidation (a 1lomerization) reactions by compound of low molecular weight can molecular oxygen that occur at position produce an equimolar concentration of C-10, which ultimately results in the an impurity resonance. Samples for NMR rupture of ring V. Excellent reviews are preferably dissolved in purified on the chemical properties and reac- and inert solvents, and the sample tions of the chlorophylls by Seely(^) tubes sealed off in a high vacuum after and more recently by Jackson (5) are thorough degassing. The manipulation available. of chlorophylls is best carried out in The central magnesium atom chelated nitrogen-atmosphere gloved boxes. Chlo- by the chlorophyll macrocycle is a reg- rophyll samples for NMR kept in air may ular rather than a transition metal ion be altered so rapidly that they can no such as is present in porphyrin-con- longer be safely used after only a few taining respiratory pigments, oxidases, hours, but NMR samples prepared from and the like. The Mg atom of the chlo- pure components in sealed tubes show no rophylls has significant electrophi1ic changes for months or even years. XH properties found to a distinctly lesser and 13C chemical shifts are given in 6, extent in the corresponding transition ppm, relative to TMS, unless otherwise element complexes. The keto C=0 group indicated. at C~9 in ring V endows the chlorophyll molecule with nucleophilic properties C. Chemical Shift Assignments that have no parallel in the porphyrins. Esterification of the propionic Despite the structural complexity of acid by phytol, a long-chain aliphatic the chlorophylls, chemical shift alcohol, makes for solubility propel— assignments are straightforward, and ties different from those of heme, indeed, more readily accomplished than which contains the free acid. The in the case of many simpler appearing chlorophylls are for all practical pur- compounds. There are many protons on poses insoluble in water and must be the chlorophyll macrocycle sufficiently studied in organic solvents. There are, isolated not to experience spin-spin of course, many other important differ- interactions sufficient to complicate ences between chlorophylls and porphy- the spectra; the methine protons, the rins, but those indicated here are per- proton at position C-10, the methyl haps the most significant in terms of groups (in Chi cO at positions la, 3a, their consequences for NMR spectros- kb, and 10b are wel1-separated and thus copy. By far the best studied chloro- appear as singlets. Where spin-spin phyll is chlorophyll a (Chi a) the interactions occur, as in the vinyl principal chlorophyll in green plants group at position 2, the protons of and blue-green algae (cyanobacteria), ring IV, and the ethyl group at posi- and most of the discussion in this tion k, the resonances are still well review will deal with this chlorophyll. separated, and their multiplicity con- tributes to the assignment. Where a B. Exper imental Aspects high-field resonance originating in a macrocycle side chain is overlaid by

Vol. 5, No. 1/2 resonances from aliphatic protons in al. (7a) suggests that the alicyclic the phytyl chain, the phytyt group can ring V has no appreciable effect on the be replaced by transesterification with macrocyclic ring current, but that the methanol. The methyl pheophorbides keto C=0 group at position 9» and the (Mg-free derivatives) and methyl chlo- addition of 2h atoms in ring IV in the rophyll ides (obtainable in si tu enzy- chlorin both reduce the ring current by matically) have simple spectra in which about 6 and 10% respectively. The sen- all of the macrocycle proton resonances sitivity of the ring current effects to are clearly visible. geometry is primarily responsible for The highly characteristic features the unusual amount of structural infor- of the chlorophyll XH NMR spectra are, mation that can be deduced from NMR however, to a considerable extent the data on chlorophyl1-nucleophile and result of interatomic induced fields chlorophyl1-chlorophylI interactions originating in the highly aromatic mac- (cf. sections I I.D.3 and h). rocycle. Such ring current effects have long been known to be important in the 1. *H NMR Chemical Shifts of Methyl *H NMR spectra of aromatic compounds, Pheophorbides and were early recognized by Becker and Bradley (6) and Abraham (7.7a) to have All of the macrocycle ring protons particular significance for porphyrin of Chi a (33 of the 72 protons in the and chlorin NMR. The ring current cal- molecule) have been assigned. The com- culations of Janson et al. (8) for oli- plete spectral assignment of Chls a and gomeric silicon and germanium phthalo- b depends to a considerable extent on cyanins have been successful in the assignment of resonances of the calculating ring current shifts in corresponding methyl pheophorbides these compounds, but this method has (chlorophylIs in which the central Mg yet to be applied to the chlorophylls. atom is replaced by 2H and the phytyl Abraham et al. (7a) have advanced a chain by a methyl group). Partial double dipole model of the macrocyclic assignment of the *H NMR of the chlor- ring current in the dehydroporphyrin phyll derivatives chlorin e6 (9). and ring of chlorophyll derivatives. This rhodochlorin dimethyl ester (9). and of X model accounts reasonably well for H the methyl pheophorbides of the chloro- chemical shift differences between cor- phylls from green photosynthetic bacte- responding protons in methyl pyropheo- ria (10) had been made prior to the phorbide a and its- porphyrin analog full assignment of the methyl pheophor- 2-vinyl-phylloerythrin methyl ester, in bide a and b chemical shifts by Closs which the additional hydrogen atoms et al. (11). A review of chlorophyll present in Ring IV of the methyl pyro- NMR work prior to 1966 describes the pheophorbide a have been removed by rationale of the chemical shift assign- oxidation. In a qualitative way ring ments of the methyl pheophorbides (12), current effects account for the unusu- and more recent reviews (13~15) cover al ly broad range of chemical shift val- subsequent developments. ues typical of the chlorophylls. The 1H Table I summarizes 1H NMR chemical NMR chemical shifts of Chi a have a shift data for the methyl pheophorbides range of 10 ppm, and the Mg-free pheo- derived from a number of important phytins a range of 12 ppm. The methine chlorophylls. The low field chemical protons in the plane of the macrocycle shifts originate from the methine are deshielded and appear at unusually bridge protons and the proton of the low field. The ring methyl, vinyl, and formyl group in methyl pheophorbide b. propionic acid protons are likewise The methine assignments for methyl deshielded to a significant extent. In pheophorbide a are based on the consid- the pheophytins (chlorophylls in which erations that the proton lies between a the Mg is replaced by 2H) the H atoms pyrrole and a pyrroline ring and attached to the pyrrole N are strongly should, therefore, be the most shielded shielded by the ring current and come methine proton (6,9)» and that the 3 into resonance several ppm above TMS. methine proton, because of its proxim- The ring current model of Abraham et ity to the ring V keto carbonyl group

Bulletin of Magnetic Resonance o

Table 1

'H NKR Chemical Sh1ftsaof Methyl Pheophorbides In C!HC1:.(12)

Methyl Methyl Methyl Methyl Methyl 2-v1nyl Pheophorbide a_ Pyropheophorbide a_ Pheophorbide b_ Bacter 1 opheophorb 1de Bacter 1 opheophorbIde cb Proton (O.O6M) (O.O6M) (O.O8M) (0.04M) (0.05M)

cx 9. 15 9.20 9.76 8.96 9.44 3 9.32 9.32 8.89 8.47 9.49 5 8.50 8.50 8.47 8.40 2a 7.85 7.98 7.75 - 7.90 2b G. 12/G.04 6.25/6 15 6.16/6.08 3. 15C 6.22/6.09 10 6.22 5. 13 6.22 6.08 5.23 8 4 .40 4.42 4.45 4.28d 4.57 7 4. 13 4.23 4. 15 4.02e n. r. '5-CHa - - 3.86 10b 3.88 - 3.95 3.84 5a 3.62 3.58 3.46 3.48 1 .96' 7d 3.57 3.58 3.62 3.57 3.60 4a 3.48 3.50 3.37 2.20 1 .67 1a 3.32 3.35 3.28 3.44 3.48 3a 3. 15 3. 13 10.58 1 .72 8a 1 .82 1 .72 1 .88 1 .79 1 .48 4b 1 .60 1 .55 1 .48 1 . 10 1 .67 N-H9 -1 .75 -1 .85 0.83 0.46 -2. 15 -0.96

a) In <5, ppm, downfield from TMS. b) A mixture, with position 4 occupied by ethyl, n-propyl, and 1-butyl, and position 5 occupied by methyl and ethyl. The a-hydroxyethyl group normally present 1n Bchl d has been converted to vinyl (32). c) The methyl group of the acetyl group at position 2. d) Includes the proton at position 3. e) Includes the proton at position 4. f) The methyl group 1n an ethyl group at position 5. g) The pyrrole nitrogen atoms. The 2 protons are not equivalent and may appear as two resonances. A minus sign Indicates a shift at higher field than TMS. is more strongly deshielded than the a the natural chlorophylls (section proton. In methyl pheophorbide b, the I I.C.3) • assignment of the a and @ protons is The low intensity multiplets near reversed on the presumption that the 4.35 ppm in all the methyl pheophor- formyl group should strongly deshield bides have been assigned to the protons the a proton so that its resonance in at position 7 and 8 (and positions 3 the b series appears at the lowest and k in Bchl a and its derivatives) on field for the methine protons. In the basis of shielding considerations methyl bacter iopheophorbide a_, all and the complicated splitting patterns three of the methine protons are posi- expected for these protons. These tioned between a pyrrole and pyrroline assignments have been confirmed by ring, and all three resonances are at a decoupling experiments that show the relatively high field. The a proton 7~proton is coupled to the methylene must be the least shielded because of protons of the propionic acid side .the acetyl C=0 group at position 2, and chain at ~2.50 ppm, and the 8-proton to the assignments of the £ and <5 reso- the high field methyl group doublet at nances are based on the arguments used ~1.80 ppm. A maximum coupling constant for the assignment in methyl pheophor- of ~2.8 Hz has been estimated for the bide a. These methine assignments are spin-spin interaction between the 7" consistent with the results of disag- and 8-protons, which is consistent only gregation titration experiments with a trans relationship between the described in section I I .D.I*. proton and the alkyl groups in ring IV. The resonances in the region 5~8 ppm The chemical shift difference between are assigned to vinyl or to other the 7" and 8-protons is surprisingly strongly deshielded substituents. The large. Originally it was attributed to vinyl group at position 2 in methyl the deshielding effect of the adjacent carbomethoxy function at position 10. pheophorbides a and b and methyl pyr- However, the difference in chemical pheophorbide a_ are easily recognized as shifts occurs also in the pyro-deriva- an AMX spin-splitting pattern, from tives of chlorophyll, in which the car- which by a standard analysis, the fol- bomethoxy group has been replaced by H. lowing coupling constants for methyl The chemical shift difference between pheophorbides a and b respectively (in the 7~ and 8-protons probably arises Hz) can be extracted: NH (x) H (A) I from the anisotropy of the macrocyclic 18.7, 18.3; MHJ[X)H(B)I 10-9. H.2; and J Uo ] ring. As the macrocyclic ring is not ! H (A) H (B) I > -*>- planar, the f- and 8-protons may occupy The C-10 proton in methyl pheophor- positions that place them at different bide a and b, is assigned to a sharp distances from the ring and so subject resonance at ~6.2 ppm, which coincides them to differing ring current desh- in many solvents with the high field ielding. Another possibility is that portion of the vinyl proton resonances. the chemical shift differences between Assignment is made on the basis of ring the ring IV protons are the result of current considerations and the proxim- substitution at the y methine position, ity of deshielding functional groups; as Abraham et al. (16) have shown that the exchange behavior of this proton the effects of methine substitution in confirms the assignment. In methyl porphyrin NMR spectra are much larger pheophorbide a and b_, there are two than can be accounted for by the protons at the C-10 position, and these resulting changes in bond anisotropies. are not magnetically equivalent, having chemical shifts that differ by about The macrocycle ring methyl groups at 0.12 ppm. These two protons yield a positions la, 3a, 5a, and 10b have highly distorted AB pattern (Ao 6 Hz, J resonances located between 3 and h ppm ~ 20 Hz) in which the two central reso- and have been assigned with considera- nances of the expected quartet are the ble certainty. In the methyl pheophor- most prominent features. The magnetic bides, the resonance of the CH3-group non-equivalence of the two proton sites on the propionic side chains can be at the tetrahedral C-10 provides impor- differentiated from the methyl group of tant information about the epimers of the carbomethoxy function at position

8 Bulletin of Magnetic Resonance l 2 10 by the synthesis of [C H3]- H methyl non-exchangeable protons, finds use in pheophorbide by transesterification of 1H NMR work. It should be pointed out the phytyl group of fully deuterated that the pheophytins are di-protonated Chi a with methanol of ordinary isto- in this strong acid, and form dica- topic composition; under the usual tions, pheoH2**. Trifluoroacetic acid transesterification conditions only the has also been employed as a solvent for propionic ester function undergoes the chlorophylls, on the unfounded reaction. The isotope hybrid methyl assumption that the central Mg atom is pheophorbide in which the position of retained. In fact, the chlorophylls the C1H3-group is independently estab- dissolved in trifluoroacetic acid lose lished makes possible an unequivocal their Mg atom and are protonated and assignment of the ester methyl groups converted to the dication of the corre- in the methyl pheophorbides, and illus- sponding pheophytin. Not only are the trates one of the ways in which fully optical properties of pheohh** remarka- deuterated chlorophylls (17) find use bly similar to those of the correspond- in NMR spectroscopy. The assignment of ing chlorophyll, but the XH chemical the remaining macrocycle methyl groups shifts of the two are also very similar is largely from the disaggregation (18). titration studies described below. Unlike the case of the Mg-containing The assignment of the high-field chlorophylls, the chemical shifts of proton resonances of the methyl pheo- the Mg-free pheophytins and pheophor- phorbides is completely straightforward bides are strongly concentration depen- and follows directly from double reso- dent. As the concentration increases, nance experiments. The elimination of 7r-7r stacking occurs to an increasing the large group of resonances from the extent (11,12), but as stacking occurs phytyl moiety greatly simplifies the with only partial overlap, selective spectra and does not significantly ring-current chemical shifts are affect the position of the macrocycle observed. The coordination interactions proton resonances. The chemical shifts between chlorophylls yield products of of the pheophytins are to a good first rather different geometry, and the approximation the sum of the methyl effects of concentration on chlorophyll pheophorbide and phytol chemical NMR spectra are considerably smaller shifts. As the capabilities of modern than for the Mg-free derivatives. NMR spectrometers have improved it has Brockmann et al. (19) have examined the become possible to see many more highly concentration dependence of the lH NMR resolved phytyl resonances. This is spectra of methyl pheophorbides poss- particularly the case when chlorophylls essing an a-hydroxyethyl group (deriva- or pheophytins of suitable adjusted tives of Bchl £). At high concentra- isotopic composition are used. In tions (>0.1 M) doubling of many of the 2H-Chl a containing 1% 1H, all the resonance lines is observed, which is methyl resonances of the phytyl moiety interpreted by Brockmann et al. to be a can be seen under 2H-decoupling, as consequence of aggregate formation. well as a number of the -CH2- reso- This conclusion is somewhat suprising, nances. These are at present unas- as aggregates produced by either ir-ir or signed, but there is no reason to sup- coordination interactions involving Mg pose that an assignment will not be form and disaggregate on a much faster forthcoming in the future. The phytyl time scale than that of the NMR meas- resonances are given in Table 2. urements. Consequently, at ambient Because of the near identity of the temperature only one set of lines has chemical shifts of the pheophytins and been observed in these systems. Pheo- the methyl pheophorbides, those of the phorbides containing a-hydroxyethyl pheophytins are not tabulated here. groups, however, appear to show line doubling, and this has been attributed A solvent sometimes employed in NMR to hindered rotation around the C-C studies because it contains no protons bond attaching the hydroxyethyl group is tr i f luoroaceti c acid~di, CF3C0.22H. to ring I. Whether hindered rotation is This is an excellent solvent for the responsible, or whether some other pheophytins, and as it is free of

Vol. 5. No. 1/2 Table 2

1H NMR Chemical Shifts3'" of Monomer Chlorophylls a, b, £1, £2 and Pyrochlorophy11 a (13)

Proton Chi ac Pyrochl ad Chi bc Chi cie Chi cif

Methine a 9-23 9.22 9.87 (9-95) (10.10) P 9-50 9.46 9.55 (9-90) (10.00) 6 8.28 8.37 8.18 (9.80) ( 9-92) 3-CHO 10.92 2-Vinyl Hx 7.92 7-99 7.• 85 8.28 8.33 6.34 HA 5.97 5-99 5..98 6.35 6. 6.04 6.06 HB 6.13 15 4-Vinyl Hx 8.33 6.32 6.04 7-Acrylic 7a 8.89 8.99 7b 6.61 6.67 10-H(2) 6.22 4.33 6.10 6.72 6.84 7 4.14* 4.21 4.15 8 4.27 4.09 4.45 la 3.28 3.22 3.22 (3-5-4)9 3a 3.25 3.16 (3^5-4)9 (3-5-4)9 4b 1.72d 1.58 n.r . 1.67 5a 3.60 3.22 3.52 (3-5-4)9 8a 1.78d 1.64 n.r. (3*.5-Mj I Ob 3-97 - 3-95 7a 2.0-2.5 -2.09 -2.35 n.r. n.r. 7b 2.0-2.5 ~2.4O ~2.35 n.r. n.r. P-l 4.41 4.38 P-2 4.89 4.97 P-3 1.42 1.45 P-4 1.74 1-75 P-CHs's 1.18 1.17 1.16 1 .12 P-CH3's O.78 0.77 0.75 0.74 0.74 0.70 0.71 0.67 0.68 0.64

a) Chemical shifts in 0, ppm relative to internal TMS. b) Chemical shifts enclosed in parentheses have been assigned from intercomparison 2 2 2 with other chlorophylls. c) In C H3C1/C H3O H (11). d) In acetone- 2 2 2 H6. e) In tetrahydrofuran- He (22,23). f) In pyridine- H5 (24,47). g) In TFA (22) .

10 Bulletin of Magnetic Resonance cause must be sought for the line doub- and 4, as does Chi a, but Chi £2 has ling still is not clear. two vinyl groups at positions 2 and 4 (21,22). Deconvolution and integration of the methine proton region can be 2. JH NMR Chemical Shifts of the Chlo- used to estimate the relative amounts rophylIs of Chi £1 and £2 in a mixture of the two (23). Chemical shifts of Chi £1 and Proton chemical shifts of Chi a, £2 are listed in Table 2. Bchl a and the important derivative BacteriochlorophylI b is present in Pyrochl a are listed in Table 2. The Rhodopseudomonas vi r idi s and a few rationale of the chlorophyll assign- other photosynthetic bacteria. This ments is very much the same as for the chlorophyll is responsible for the methyl pheophorbides. As indicated extreme long wavelength light absorp- below, the XH NMR spectra of the chlo- tion in these organisms. The most rophylls are strongly solvent depen- striking feature of the Bchl b struc- dent, and the relationship between the ture is the ethylidene side chain at spectra in polar (nucleophi1ic) and position 4, which replaces the ethyl non-nucleophi1ic solvents provided group present in Bchl a. The main dif- valuable information for the assignment ferences in the XH NMR of Bchl b com- of the resonances observed in nucleo- pared to Bchl a are the resonances of phi 1 ic solvents. the ring II protons (13,24). Both the It should be noted that phytol is by 3a and 4a protons give rise to double no means the universal esterifying doublets (J1 = 2 Hz, J2 =7 Hz) at low alcohol in the chlorophylls. While all field (6 = 4.93 and 6.84 ppm). Double samples of Chi a and b so far examined resonance experiments show them to be appear to be esterified primarily by coupled to each other (J = 2 Hz) and to phytol, Bchl a derived from Rhodospi- a high field methyl group each (J '•- 7 ri1lum rubrum is esterified principally Hz). The double bond in the ethylidene by geranyl geraniol (a C20 alcohol with group shifts the (3-proton resonance to k double bonds) (20). It has long been lower field. All other resonances are known that the chlorophylls from green identical with those ot Bchl _a (Table photosynthetic bacteria (Bchl c_) con- 3). tain farnesol (a C15 alcohol with 3 Green photosynthetic bacteria con- double bonds). The presence of addi- tain very complicated mixtures of chlo- tional vinylic protons or methyl groups rophylls whose exact structures are in farnesol or geranyl geraniol pro- still under active investigation. duces additional olefinic resonances in Referred to in early publications as the region 5 k~S ppm and this possibil- "chlorobium" chlorophylls because of ity must be kept in mind in the analy- their isolation from Chlorobium species sis of pheophytin and chlorophyll spec- (25) they are more often called bacter- tra. iochlorophylIs £, d, and e. These chlo- Some more recently characterized rophylls are unique among all natural chlorophylls merit comment. Chloro- chlorophylls in that each appears to be phylls £1 and C2 are auxilliary acces- a mixture of various homologs. Thus, sory pigments in marine diatoms and Bchl c, d, and e are each families of brown algae. These chlorophylls are chlorophylls containing homologous closely related to each other and to alkyl groups at positions 4 and 5. Bchl Chi a. Unlike Chi a, however, they are £ and e in addition contain a CH3-group porphyrins, not chlorins, although an at the 6 methine position. Recently, intact ring V is present in both. Chi another series of Chlorobium chloro- ci and £2 are both free acids, and lack phylls isolated from Chlorobium phaeo- an esterifying alcohol at the position vi roides has been investigated and 7 side chain. The side-chain substit- described by Brockmann and co-workers uent at position 7 is a transacrylic (26,27). This family of chlorophylls acid group (AX pattern). Chi £1 and £2 (designated Bchl e) contains a formyl differ from each other in that £1 has a group (established by the appropriate vinyl and ethyl group at positions 2 CHO resonances in both the *H and 13C

Vol. 5, No. 1/2 11 Table 3

'H NMR Chemical Shiftsa of Bacter1ochlorophyl1s a and b, and the Methyl Pheophorbides of Bacter 1ochl orophyl 1 s c, d_, and e_

Methyl Methyl Methyl Bchl ab Bchl bb Bacteriopheophorbide cc Bacteriopheophorbide d Bacteriopheophorbide ee Proton (0.06M) (0.06M) (0.08M) (0.04M) (0.05M)

9 . 23b 9.41 9.90 9.57 10.58 f-J 9.5O 8 .93 9.41 9.28 9.42 0 8.28 8. 39 - 8.45 - 10 6.44 6.43 5. 17 5.04 5.20 3 4. 10 4.93 (dd) - - - 4 3.86 - - - - 7 4 . 10 4 . 10 4. 14 4. 13 n. r . 8 4.21 4.21 4.55 4.36 4.58 2a - - 6.47 (q) 6.31 (q) 6.56 4a ~2.5 6.84 (dd) 3.68 (q) 1 .68 (t) 1 .72 1a (3.33) (3.34) 3.48 3.38 3.53 2b (3.OO) (2.99) 2.12 (d) 2.08 2.15 3a 1.58 (d) 1.66 (d) 3.26 3.19 1 1 .07 4b . ' . n. r . 2.01 (d) 1.68 (t) - 1 .20 5a (3.44) (3.45) 3.61 3.51 4 .01 8a 1.41 (d) 1.41 (d) 1.41 (d) 1 .75 1.51 10b (3.66) (d) (3.66) • - - - 5-CHa - - 3.85 - 3.86 7a ~2 . 5 ~2.5 - - L. 7b ~2.5 -2.5 - - - 7d - - 3.62

;t i n 3.58 3.62 5b 1 .92 o a) Chemical shifts in 6, ppm relative to internal TMS. Chemical shifts in parantheses are assigned by interconversion ~h with other chlorophylls. b) In pyridine-'Hi, (24) . c) In C! HCl.i. This sample contained methyl groups at positions 3 and 5, and an ethyl group at posit1on 4 (30). d) In C * HC 1 :.. This sample contains an ethyl group at position 4 and a Q methyl group at position 5 (31). e) T n C ' HC 1 :.. This sample contained a mixture of ethyl, n-propyl or Isobutyl side ne t i chains at posi tion 4 (3). d) Includes the proton at position 3. c Resonanc e NMR spectra) and, thus, has the same of the methine chemical shifts of relationship to Bchl £ and d as does methyl pheophorbide a. Chi b to Chi a. All of the Chiorobium Sanders and co-workers (33.3*0 have chlorophylls have an (a-hydroxy- reported spin-lattice relaxation times ethyl)-substituent at position 2 char- (Ti), nuclear Overhauser enhancements acterized by a low-field quadruplet at (NOE), and long-range coupling con- 6.1-6.6 ppm and a high-field doublet. stants for the chlorophylls. The T1 These chlorophylls all lack a 10 CO2CH3 values for the methyl protons depend group and are, thus, pyrochlorophyl1 largely on distance from the macrocycle derivatives. The AB double doublet and steric crowding, but the T1 values expected for the IO-CH2 protons is for the methine protons are dependent often only poorly resolved (10). The on the substitution pattern (Table k). predominant esterifying alcohol in Bchl In the absence of any information on c (the chlorobium-650 of Holt (28)) is the T1 error limits it is difficult to farnesol, but Strouse et al. (29) have judge the usefulness of relaxation shown that small amounts of at least times that fall in a narrow range for five other esterifying alcohols are making chemical shift assignments. San- present. Risch et al. (30) have like- ders et al. (33) attribute line-width wise found that the chlorophylls from variations in the chlorophyll spectra Chloroflexus aurantiacus contains stea- to unresolved long-range acyclic ryl (C18H37), phytyl (C20H39) , and ger- couplings. The assignments of Sanders anyl geranyl (C20H33), and not farnesol et al. made on the basis of T1, NOE, as the esterifying alcohols. and long-range coupling effects agree, To simplify the application of NMR however, in all particulars with previ- and mass spectroscopy, it has been cus- ous chemical shift assignments made tomary to eliminate the long-chain from ring-current and disaggregation aliphatic alcohol by transesterifyica- considerations. tion with methanol, a procedure during which the Mg is lost. The methyl pheo- 3. Chlorophyll Related Structures phorbides of Bchl c, d, and e are more easily separated by chromatography than A number of structures related to the chlorophylls themselves (31). Con- the chlorophylls have been character- sequently, all of the available NMR ized by lH NMR. These include the epim- data on Bchl c_, d, and e is for the ers, enol, and the Krasnovskii photore- methyl bacteriopheophorbides. Selected duction product of Chi a, and this has data for some of the numerous homo logs contributed significantly to the clari- of the bacteriomethylpheophorbides are fication of some longstanding problems given in Table 3- in chlorophyll chemistry. The methyl pheophorbides of Bchl £ It has long been known that Chi a and d have been used by Trowitzch (32) and b, in the course of column chroma- to assign the methine chemical shifts tography on sugar, are accompanied by in methyl pheophorbide a_ and methyl small, faster- running satellite bands, 1 pyropheophorbide a. In the Bchl £ and d designated a' and b (35)• The two sub- derivatives, the a and 3 protons have stances are easily interconvertible, distinctly different neighbors, unlike and it was suggested by Strain (35) 1 the situation in methyl pheophorbide a that a and a were diastereomers, epim- or methyl pyropheophorbide a, and eric at carbon-10. Experimental evi- assignment of the a and (3 protons is dence for this interpretation was pro- facilitated. Conversion of the hydroxy- vided by *H NMR studies (36)., which ethyl group to vinyl converts methyl showed that the diastereotopic C-10 bacteriopheophorbide d to methyl pyro- protons of a and a/ had chemical shifts pheophorbide a, and the spectral closely resembling those of the two changes support the original assigment C-10 protons of pyrochlorphyll a_.

Vol. 5, No. 1/2 13 Table k

Spin-Lattice Relaxation Times (s) for Chlorophylls

Protons Chi a Chi b Methyl Methyl Bchl a Chlorophy11ide a ChlorophylVide b

a 1.0 0.8 1.3 1.5 1.0 P 0.9 0.6 1.1 1.1 0.7 0 1.0 0.8 1-3 1.3 0.6 10 ).ha 1.2 l.7 1.6 0.8 7 0.7 - - 0.5c 8 0.6 0.6 - o.5c la 0.7 0.6 0.9 0.9 0.5 2b - - 1.3d 3a 0.6 - - 0.3 kb 0.7 0.6 0.08 0.7 5a 0.7 0.7 1.2 0.5 8a 0.4 0.5 0.5 0.4 10b 1.0 0.8 1-3 0.8 7d - - 1.8 1.8

a) For Chi a1, Ti = 2.0 s. b) For methyl chlorophy 1 ide a1, Ti = 2.0 s. c) Estimated by null point because of signal over 1ap. d) The methyl group of the acetyl function at position 2.

In Pyrochl a, the two protons at C-TO protons (and other resonances as well) are in magnetically non-equivalent in a' are easily resolved from those of positions (37) and mixtures of Chi a_ a, and can often be clearly distin- and a_' have peaks in their 1H NMR spec- guished in the 1H NMR spectrum of an tra that can be assigned to analogous equilibrium mixture of Chi a_ and a_'. In diastereomeric C-10 protons. Recently, an XH NMR study, Ellsworth and Storm an alternative interpretation for the (41) have shown that the Mg-free methyl structure of Chi a' was revived (38) , pheophorbide a_' is much less prone to which claimed that Chi a1 is the enol isomerization than is Chi a'. In chlo- form of Chi a. To resolve the situ- roform solution at room temperature, ation, Chi a and Chi a_' were separated methyl 10-epipheophorbide a appears to chromatographically at 0°C, and XH NMR be stable indefinitely. The difference spectra recorded on the eluted compo- in the rates of epimerization is nents at low temperatures, where inter- attributed to conformational differ- conversion between a and a_' is very ences between the chlorophyll and pheo- slow. Chi a and a1 are clearly seen to phorbide. As in the case of Chi a and have C-10 protons with different chemi- a1, the methyl pheophorbides a and a' cal shifts (Table 5). thus disproving have significantly different chemical the enol hypothesis and establishing shifts for the methine, C-10, carbome- Chi a' as the epimer of Chi a (i+0) . In thoxy, and la, 3a, and 5a methyl pro- further work, Hynninen and Sievers tons. (40a) deduced from additional 1H NMR Ring V in the chlorophylls contains data that conformational changes (puck- a (3-keto ester function and is there- ering) of the whole macrocycle occurs fore prone to enolization. In solution, with epimerization at C-10. It is the keto/enol equilibrium in all of the interesting to note that the methine chlorophylls is strongly displaced

Bulletin of Magnetic Resonance toward the keto form, and only a small, Chi a (48). The structure of the photo- stationary concentration of enol reduction product remains elusive, how- appears to be present. The enol has ever, despite much study (49). XH NMR been implicated in the Molisch phase studies have now made possible a struc- test, which establishes the integrity ture assignment to the photo-product of ring V (42) as an intermediate in (50). The photoreduction of Chi a is the allomerization reactions of chloro- carried out with 1H2S or 2HaS directly 2 phyll (4,5). and in hydrogen exchange in a sealed NMR tube. When H2S is at position C-10 (43). Interest in the used as the reductant, the already sim- enol remains keen, for many models have ple NMR spectrum of the photo-product been advanced involving enol participa- is even further simplified. From the XH tion in photosynthetic oxygen evolution NMR spectra it is immediately evident (44) and in the primary events of pho- that the photereduction of Chi a tosynthesis (45) . results in the loss of the ring cur- Peripheral complexes are formed from rent, i.e., the conjugation in the mac- pheophytins or methyl pheophorbides and rocycle is disrupted. Most of the lower Mg 2 (46). Peripheral complex formation field resonances of Chi a are shifted with Mg-containing chlorophylls does to substantially higher field, while not occur to a significant extent. 1H the signals originating in the phytyl NMR shows that peripheral complexation moiety remain substantially unchanged. occurs with the enol form of the The upfield sh*ift is most pronounced in 3-ketoester system of ring V. The C-10 the resonances of protons closest to proton is no longer to be seen in the the macrocycle. The upfield shifts are XH NMR spectrum of the peripheral com- of the order of 1.0-1.7 ppm for the plex. The ring-current induced shifts vinyl and ring methyl protons, and in these complexes are smaller than in about 6 ppm for the (3 and 5 methine the free pheophorbides. The IO-COOCH3 protons. As the integrated area of becomes more or less coplanar with the these two resonances indicates the macrocyclic systems, and the incremen- presence of 2 protons, it is concluded tal low-field shift is unusually low, that the Krasnovskii reaction product which is a direct result of the move- is P,5-dihydro-chlorophyll a. lH NMR ment of this group into a deshielding also establishes that the reversal of region of the ring current. A compari- the photoreaction in the dark restores son of the chemical shifts of methyl the original Chi a. The increased sen- pheophorbide a and its peripheral Mg sitivity of modern NMR spectrometers complex is shown in Table 6. makes it possible to study the effect The enols of Chi a, Pheo a, and of light irradiation on chlorophyll methyl pheophorbide a have been trapped solutions sufficiently low in concen- as the tetramethylsilyl ethers (40). tration to permit photochemical inves- The silyated enol of Chi a is labile, tigations in the spectrometer probe, and easily reverts to the original Chi and such investigations very likely a, or is converted to the silylated will open a new chapter in chlorophyll enol of Pheo a. The XH NMR chemical photochemistry. shifts of methyl pheophorbide a and its enol trimethysilylether are compared in 4. 13C NMR Table 7- The largest changes are observed in the chemical shifts of the All 55 carbon atoms in Chi a have methine protons, which again implies a had their 13C NMR chemical shifts large decrease in the ring current in assigned. General studies of the 13C the stabilized enol. NMR spectra of chlor ins have been The Krasnovskii photoreduction of reported by Lincoln et al. (51)» and Chi a was the first and is possibly the Smith and Unsworth (52), and Chi a most widely studied photoreaction of itself has been studied by a number of the chlorophylls. Chi a dissolved in research groups (53~59) • Assignment of pyridine can be reversibly reduced in the quarternary carbon atoms was car- light by ascorbic acid to a pink photo- ried out by Boxer et al. (56), while product, which in the dark reverts to Goodman et al. (60) have assigned all

Vol. 5, No. 1/2 15 Table 5

NMR Chemical 5hiftsa of Methine and C-10 Protons in Chlorophyll a and a1 (39)

In Pyridine-2Hs In Acetone-2He

Proton Chi a Chi a1 Chi a Chi a1

a 9.1»2 9-39 9.07 9.04 (3 9-58 9.56 9.40 9.37 6 8.28 8.26 8.26 8.22 . C-10 6.44 6.30 5-99 5-87

a) 8, ppm, relative to internal hexamethyl disiloxane.

Table 6

NMR Chemical Shifts3 of Peripheral Mg Complex of Methyl Pheophorbide a (47)

Proton Per ipheral Methyl Mg Complex" Pheophorbide ac

a 8.83 9.47 (3 9.01 9-75 6 8.00 8.71 Vinyl Hx 7-77 8.08 HA 6.06 6.23 6.05 HB 5-87 10 - 6.61 7 lt.65 4.29 8 4.10 4.42 10b 3.83 3.76 7b 3.38 3.52 5a 3-11 3-42 3a 2-95 3-21 la 2.83 3.08 8a 1.73, 1.66 4a 3.29 3.54 4b 1.39 1-53 N-H 2.44 +0.74 N-H 2.04 -1.48

a) Chemical shifts in 0, ppm relative to internal TMS. b) Recorded on a solution of pheophytin or pheophorbide a (7*10 3 M) in a saturated solution of Mg(C104)2 in pyridine-2Hs. c) Spectrum of the free methyl pheophorbide a regenerated by addition of 10 /nl of 'HhO.

16 Bulletin of Magnetic Resonance Table 7

*H NMR Chemical Shiftsa of Methyl Pheophorbide a and the Trimethylsilylether of the Enol of Methyl Pheophorbide a_ (40)

Proton Methyl Pheophorbide a Trimethylsilyl Ether of the Enolc

a 9-25 8.19 3 9,33 8.24 6 8. 18 7..03 Vinyl 7.• 71 7. 19 5-•98 5-.64 5-• 78 5.46 10 6..20 7 4. 16 .66 8 3-.95 .22 10b 3-• 37 .60 7d 3-.26 07 5a 3. 18 73 4a 3. 13 98 la 2.• 91 .46 3a 2.• 90 37 7a 2..29 .81 7b 2,.06 73 8a 1 .5.1 27 4b 1 .42 .22 N-H 0.79 .26 -1.38 2.14

a) Chemical shifts in 6, ppm. b) In benzene-2He. Chemical shifts relative to internal hexamethyldisiloxane. c) In ben2ene-2H6. Chemica shifts relative to the trimethylsi1yl group of the compound.

of the carbon atom resonances in the and Lotjonen and Hynninen (62a). The phytyl moiety. Argonne studies used revisions have been incorporated into Chi a enriched to 15-20% 13C and Matwi- Table 8. Insertion of Mg into methyl yoff et al. (58,59) used Chi a of 90% pheophorbide a produces downfield 13C enrichment, in both instances pre- shifts of carbon resonances in rings I pared by biosynthesis with 13C02 of the and III, and upfield shifts for the appropriate isotopic composition. resonances associated with the carbon Chemical shift assignments for the atoms in rings II and IV. Coupling con- carbon atoms of methyl pheophorbide a, stants (JI3£_H) f°r carbon atoms methyl pyropheophorbide a, and Chi a bearing protons are listed in Table 9- are listed in Table 8. The original 13C NMR spectra recorded on Chi a and b assignment by Boxer et al. (56) of C-6, containing 90% 13C have been made it C-l6, and C-17 have been revised by possible to extract a number of 13C-13C Smith et al. (6l) , Wray et al. (62), coupling constants, which are listed in

Vol.. 5, No. 1/2 17 Table 8

13C Chemical Shifts (<5, ppm)a of Monomeric Chlorophyll a, Methyl Pheophorbide a^, and Methyl Pyropheophorbide a

Carbon No. Chiorophyl1 ac Methyl Methyl Pheophorbide ad Pyropheophorbide a&

la 14.9 or 15.0 11.8 11.8 2a 133.*• 128.3 128.5 2b 121.2 121.8 121.6 3a 13.5 10.7 10.8 4a 22.2 19.0 19.1 4b 20.2 17.1 17.2 5a 14.9 or 15.0 11.8 11.8 7a 33-6 31.0 51.4 7b 32.6 29.8 31.0 7c 175.1 172.6 29.8 7d - 51.4 172.8 7 53-3 51.0 51.4 8 51.8 49.9 49.7 8a 26.1 22.8 22.9 9 191.9 189.0 195-2 10 68.2 64.5 47.8 10a 173-1 168.9 - 10b 54.3 52.6 - a 103.0 96.4 96.4 3 110.1 103.6 103.2 y 108.4 104.8 105.4 6 95.2 92.6 92.4 l 137.6 or 136.4 131.1 130.7 2 141 .7 135.7 135.1 3 137.6 or 136.4 135-3 135-2 4 146.6 144.2 144.0 5 137.6 or 136.4 128.3 127.4 6 164.5 (13C.9)9 160.5 (128.3) 129.7 11 156.8 141.3 140.7 12 150.7 135.3 135.3 13 154.1 155.0 154.1 14 148.7 150.7 (149-7) 15 150.6 137.2 136-9 16 158.6 (162.4) 149.0 148.2 17 -174.5 (156.3)9 172.6 (160.5) 159.5 18 170.2 171.4 170.4 P-lf 63.8 P-2 122.1 P-3 144.7 P-3a 18.5 P-4 42.4 P-5 27-7 P-6 39.4 P-7 35.4 P-7a 22.3 P-8 40.1

18 Bulletin of Magnetic Resonance Table 8 (Continued)

13C Chemical Shifts (5, ppm)a of Monomeric Chlorophyll a, Methyl Pheophorbide a, and Methyl Pyropheophorbide a

Carbon No.b Chlorophy11 ac Methyl Methyl Pheophorbide 3d Pyropheophorbide ae

P-9 27-3 P-10 40.1 P-ll 35-5 P-lla 22.3 P-12 40.1 P-13 27-7 P-14 42.1 P-15 30.7 P-15a 25.2 P-16 25-2

a) Relative to hexamethyldisilane for chlorophyll a; relative to tet- ramethylsilane (TMS) for methyl pheophorbide £. b) See Figure 1 for numbering, c) In benzene/tetrahydrofuran solution, d) In C2HCl3 solution (14). The values in parentheses are from the reassignment of Smith et al. (61). e) References 52, 61. f) Phytyl carbon assignments are based on those of Goodman et al. (60) and are relative to internal TMS. g) Lotjonen and Hynninen (62a) have revised the assignments for carbon atoms 6, 16, and 17. Note that their 13C chemical shifts are relative to TMS and for acetone-d6 solutions.

Table 10. 5. 15N and 2H NMR Smith et al. (6l), in connection with efforts to resolve the complicated Full assignment of the 1SN chemical questions surrounding the structure of shifts in Pheo a and Chi a have been the Bchis j:, have collected extensive made by Boxer et al. (56) (Table 11). 13C NMR data on the methyl pheophor- The Chi a and Pheo a derived from it bides derived from the Bchls c, and contained 95% 15N, incorporated by have presented more difinitive 13C NMR biosynthesis. The 15N spectrum of Pheo assignments for methyl pheophorbide a a was recorded directly, but for Chi a_ and methyl pyropheophorbide a. Assign- the 15N relaxation times were so long ment is greatly facilitated by the use as to preclude direct observation of 13 of chlorophylls containing . 15-20% C. the spectra, and the l5N spectral For Chi a and b, this is not difficult parameters were obtained indirectly by to accomplish, but for the chlorophy11s INDOR. Long-range coupling between 15N 13 from photosynthetic bacteria, C and the methine protons was observed in enrichment is a rather complicated the lH NMR spectra of both compounds, task, which up to now has been fully as well as the expected 15N coupling solved only for Bchl a. Prospects for with the inner protons in Pheo a. Anal- obtaining the other bacterial chloro- 15 13 ysis of the N NMR spectrum of Pheo a phylls sufficiently enriched in C to also yielded all of the 15N-15N make a full assignment possible do, coupling constants (Table 12). however, appear to be good. 2H NMR spectra (at 15.4 MHz) have been reported by Dougherty et al• (63)

Vol. 5, No. 1/2 Table 9

Coupling Constants Ji3c_i^ for Chlorophyll a and Methyl Pheophorbide aa

b Ji3C_xH (Hz)

Carbon atom Chlorophyll a Methyl pheophorbide a

Vinyl C-2a 150 155 C-2b 158 160 C-Phy-2 152 -

Methi ne a 150 . 155 3 148 155 6 152 157

Aliphatic -CH C-7 - 129 C-8 130 -130 C-10 132 136

-CH2 C-4a - 125 C-7a - 130 C-7b - 126

-CH3 C-Phy-1 153 - C-la 128 129 C-3a 125 126 C-4b 160 C-5a 128 129 C-8a - 125 C-lOb 148 148

a) Table from Janson and Katz (53) • b) ±2.5 Hz.

ithyl pheophorbide a-d35 and Chi C. Applicat ions of NMR a-d72. The line widths of 2 to 7 Hz for methyl pheophorbide a-d35 were broa- Many applications of NMR to struc- dened by quadrupole relaxation, but ture determination and conformation were still sufficently narrow to permit studies have been made to the chloro- assignment of many of the 2H reso- phylls. In addition, NMR has been nances . largely responsible for major advances in the understanding of chlorophyll behavior, and has been particularly

20 Bulletin of Magnetic Resonance Table 10

Some Coupling Constants Jur-i for Chlorophylls a and b°>b

Carbon atom Chlorophyl1 a Chlorophyll b

1-la 44 44 2a-2b 68 68 3-3a 45 50 4-4a 42 42 4a-4b 34 34 5~5a 44 44 7b-7c 55 55 8-8a 46 46 10-10a 58 58

a) In Hz. b) Table from Matwiyoff and Burnham (59)

Table 11

15N Chemical Shifts for Chlorophyll a and Pheophytin aa'b

Nitrogen Chlorophyll ac Pheophytin ad

N-l 163.6 102.5 N-2 183-5 218.5 N-3 166.4 110.9 N-4 224.0 272.8

a) Table from Boxer et al. (56). b) In ppm relative to external 15NH.iCl in 2 N HC1. c) In acetone-2He. d) In C2HCls.

effective in defining the nature of was for long a subject of speculation. chlorophyll-chlorophyll and chloro- Such an hypothesis implies exchangeable phyll -nucleoph Me interactions. Some of hydrogen in either the ground state or the more significant applications of excited states of chlorophyll. XH NMR NMR in chlorophyll chemistry are now has successfully addressed both of cons idered. these problems (64,65). Both the C-10 and the o-methine protons are readily exchangeable in Chi a, Chi b, and Bchl 1. Exchangeable Hydrogen in Chlorophyll a. With the hydroxyl group of methanol in pyridine, hydrogen exchange at C-10 A possible role for chlorophyll in is rapid, whereas exchange at all meth- the light conversion step in photo- ine carbon atoms is much slower. synthesis as a cyclic hydrogen donor Exchange at the 6 position is strongly

Vol. 5, No. 1/2 21 influenced by the presence of Mg. constants of j_n vi tro Chi a and Bchl Removal of the Mg reduces the exchange a*' and in vivo P700*" and P865*' by rate at the methine bridge positions to ENDOR spectroscopy has provided consid- very low values. That the "extra" erable support for the original conclu- hydrogen atoms in ring IV do not func- sions (67). ENDOR spectroscopy, how- tion as reducing agents in a cyclic ever, is not without its problems, and process in photosynthesis is demon- the chemical manipulations required for strated by the failure to observe the preparation of selectively deuter- hydrogen exchange when green algae ated chlorophyll derivatves required grown in 99-5% 2H20 are transferred to for ENDOR assignments are not trivial 1H20 and allowed to conduct photo- (68). Sanders and Waterton have, synthesis. Chlorophyll extracted from therefore, undertaken the determination these organisms is found to contain no of the hyperfine coupling constants of ,XH by 1H NMR, arguing against a role the chlorophyll cations by NMR line for exchangeable hydrogen at the C~7, broadening in the fast exchange limit C-8, and o-methine positions in photo- (69,70). NMR in principle is a far synthes i s (65) • superior method for determination of the hyperfine coupling constants as assignment follows immediately from the chemical shift assignment, and no chem- 2. NMR of Paramagnetic Chlorophyll ical manipulations are required. Only Speci es relative hyperfine coupling constants can be deduced by NMR, and a reliable The hypothesis that the primary value from ENDOR of at least a few of electron donor in photosynthesis is a .the coupling constants is required for special pair of chlorophyll molecules conversion to absolute values. Water- was originally based on EPR lineshape ton and Sanders (70) find that NMR analysis of the cation species that results with Chi a* and Bchl a* agree well with ENDOR, but for Chi b*', Table 12 agreement is poor. The experiments of Waterton and Sanders were carried out 15 N Coupling Constants at room temperature or above to insure a in Pheophytin a the chlorophyll species were in fast exchange. The Chi a_* cation has a Coupling Constant Value (Hz) half-life at room temperature of about 20 min. and at 310 K about 5 min. The variation in Chi a_* concentration during the experiment may, thus, compli- ...5N-HI 98 cate interpretation of the line broadening data. 5N-methine H 3-0 Brereton and Sanders (70a) have 2 Jl 5Ni-15N? 2.0 studied the radical anion of bacter- iochlorophy11 a by observing differen- 5N -15N 5-7 tial electron transfer line broadening 2 3 in the 1H NMR spectrum of diamagnetic 2 Bchl a in the presence of a small Jl •N,-»N4 l.if amount of chemically produced anion, 2Jl 2-5 formed by reaction with sodium sulfide. From the observed line broadenings, it was concluded that the protons at the a) Table from Boxer et al. (56) a,3.T,7»^»5a»la,2b,3a. and 8a positions have significant hyperfine interactions with the unpaired electron in Bchl a" , but no hyperfine coupling constants were calculated. In these experiments, remains after electron transfer (66). A no chemical evidence is presented that comparison of the hyperfine coupling unequivocally establishes the chemistry

22 Bulletin of Magnetic Resonance of the reaction of Bchl a with S" paramagnetic species appear to be (see reference 50). involved in the oxygen-generating side Closs and Sitzmann (71) have suc- of photosynthesis. Wydrzynski and co- cessfully determined the hyperfine workers (73, Jk, 75) have initiated coupling constants of radical cations research in which the effects of param- of chlorophylls and derivatives by agnetic species on the lH and 170 time-resolved CiDNP (chemically induced relaxation times of water are being dynamic nuclear polarization) studies. explored in an effort to clarify the In these experiments, polarized *H NMR oxygen-evolving apparatus in green spectra are recorded on a chlorophyll- plants. Proton and 170 relaxation benzoquinone system in which electron rates (1/Ti and I/T2) have been meas- transfer is induced by a nanosecond ured in chloroplast preparations, with laser light pulse. The hyperfine results that suggest that manganese in coupling constants determined by this a mixture of oxidation states is nor- procedure are relative, and require at mally present in dark adapted chloro- least one ENDOR-determined value for plasts (74). The relaxation rates for normalization; the absolute hfc values XH and 170 are for the most part deter- obtained by CIDNP (after normalization) mined by loosely bound Mn present in agree well with the ENDOR values. This the chloroplast membranes; it is esti- procedure should have wide applicabil- mated that from one-third to one-fourth ity. It would be desirable, however, of the loosely bound Mn is present in because of the importance of these dark-adapted chloroplasts as Mn (II), studies, to explore in more detail the the remainder being in higher oxidation chemistry of the reaction used. Benzo- states. The Mn appears to be located in quinone can oxidize Chi a by a dark the interior of the photosynthetic mem- reaction, and nucleophilic attack on branes (73)- Experiments have also been the cation can form alteration prod- carried out in which the XH spin-spin ucts. The presence of methanol in (transverse) relaxation rate of chloro- these experiments raises the possibil- plast suspensions has been measured ity that 10-methoxy Chi a could be after each of a series of 2.1* jusec forming in the system during the course light flashes. The sequence of relaxa- of the experiment, which would conceiv- tion rates oscillates and has a maximum ably complicate the interpretation of value after every fourth flash. This the observations. has been interpreted to indicate that manganese participates in the charge Selective line broadening by light accumulation process during oxygen evo- irradiation has been shown by Boxer and lution (75). However, the interpreta- Closs (71a) to occur in light-excited tion of the experiments of Wydrzynski molecules in an NMR probe, and they et al. (75) has been questioned by Rob- have developed a method that success- inson et al. (75a). These investiga- fully extracts information on spin dis- tors found that the changes in the pro- tribution in photo-excited triplet ton relaxation rate can be abolished by states from high resolution *H NMR 1 removal of Mn(ll) from the chloroplasts data. Some preliminary H NMR data on with appropriate chelating agents with- the relative hyperfine coupling con- out affecting the evolution of O2. The stants in light-excited methyl chloro- manganese that is involved in the phyll ide a have been reported (72). relaxation phenomena thus appears not Spin distribution in the Chi a* cation to be the manganese involved in the free radical appears to be considerably 3 literature evolution of O2 during pho- different from that of Chl a. Rela- tosynthesis. In any event, the appli- tively little spin density is to be cation of NMR techniques would appear found at the methine positions in doub- to be of considerable promise for the let state Chi a* , whereas the methine study of the oxygen side of photo- positions have the highest spin densi- synthesis. ties in the chlorophyll triplet. Paramagnetic chlorophyll species are formed in the primary light conversion 3. Chlorophyl1-Nucleophile Interactions event in photosynthesis, and other

Vol. 5, No. 1/2 23 Chlorophyll NMR spectra are remarkably solvent dependent, being very different in nucleophilic and non-nucleophi1ic solvents. This solvent dependence was early recognized to be a consequence of nucleophilic interactions at the central Mg atom of the chlorophylls (11,76). All lines of spectroscopic investigation support the view that Mg with coordination number k in chlorophyll is coordinative!y unsaturated, and that there is in consequence a driving force to acquire •electron donor functions (i.e., lone pair electrons on oxygen, nitrogen, or sulfur) in one or both of the Mg axial positions. Chlorophylls dissolved in typical nucleophilic, polar (Lewis base) solvents such as acetone, diethyl ether, pyridine, tetrahydrofuran and the like occur as monomeric chlorophyll with one or two molecules of solvent, depending on basicity, in the axial positions of the Mg. In weak donor sol- 4.50 vents such as acetone or diethyl ether 5 10 15 20 25 the Mg occurs largely with coordination Mole ratio CH,OH/d- chlorophyll a number 5. and in more basic solvents such as pyridine, the coordination number of the Mg approaches 6. Figure 2. Chlorophyll a-methanol inter- The coordination interaction at Mg action in carbon tetrachloride solu- positions the ligand in the center of tion. Chemical shifts of CHa(O) and the chlorophyll macrocycle , where it C-10 (A) protons as a function of 2 is subject to the full force of the CH3/ H-Chl a (0.064 M) ratio. Addi- ring current. Chlorophyll is in effect tional C-10 points (o) are derived from a natural NMR shift reagent. The chemi- a methanoi titration of ordinary Chi a_. cal shifts of the protons of a ligand The solid lines are calculated curves. bound to Mg will therefore experience an upfield ring current shift to an extent determined by the distance of a particular proton from the center and plane of the chlorophyll macrocycle. methanol as measured by coordination to Katz et al. (77) have made a quanti- the Mg atom of Chi a. tative study of chlorophyl1-nucleophile Quantitative observations have been interactions by observing the ring cur- made on coordination interactions rent effect on proton chemical shifts between chlorophyll and various com- of the ligands bound to chlorophyll. pounds present in thylakoids and likely The use of fully deuterated chlorophyll to be near neighbors of chlorophyll simplifies interpretation of the spec- (77) • As expected, (3-carotene does not tra. Pentacoordinate Mg (II) appears to appear to experience any interaction dominate the equilibrium with aliphatic with chlorophyll that results in plac- alcohols, the equilibrium constant for ing any of its protons near the chloro- the formation of Chi a^CHsOH in CCU phyll macrocycle. The XH NMR spectrum solution being K1 = 56 1 mol"1 (Figure of (3-carotene in C2HCl3 solution is the 2). Georghe et al. (78) have studied same in the presence or absence of the chlorophyll-water interaction by *H 2H-Chl a. With lutein, a dihy- NMR. Water as a nucleophile is observed droxy-3-carotene, the situation is very to have about the same base strength as different, for there are major

Bulletin of Magnetic Resonance differences in the *H NMR of lutein 3-methine protons. The shift in the when 2H-Chl a is present. These differ- phytyl may then be due to displacement ences are consistent with the coordina- of the phytyl chain from the diamag- tion of the hydroxy] groups to the Mg netic zone of the macrocycle. of Chi a, which results in a marked difference in the magnetic environment k. Chlorophyll-Chlorophyll Interactions of the ring methyl groups of the lutein. Similar changes are observed XH NMR spectra of Chi a in non-nu- with other carotenoids carrying nucleo- cleophilic solvents are very different philic groups. Another important class from those in nucleophilic solvents of chloroplast components are the galactolipids, sulfolipids, and phos- pholipids. These compounds might reasonably be expected to coordinate to Mg by way of ester C=0, -OH, or -SO3H functions present in these molecules. However, in C2HCl3 solution, plant lip- ids, particularly the sulfolipid char- acteristically present in green plant photosynthetic membranes, show only a weak tendency to compete for coordination to Mg. This may be due to the presence in chloroform solution of both the chlorophyll and the sulfolipid as inverted micelles, in which the polar regions of both substances are buried in the center of the micelle. With current interest in the photosynthetic membrane, further studies of chlorophyll-lipid interaction would appear to merit attention. Larry and VanWinkle (79) have made an XH NMR study of the interactions of Chi a and b with sym-trinitrobenzene. The interactions were studied in chloroform solution containing methanol so 10.0 6.0 4.0 8,ppm that the chemical shift changes must be attributed to generalized t:—n forces rather than to coordination interac- Figure 3. XH NMR spectra of chlorophyll tions at Mg. At a molar ratio of 1:1, a_ in nucleophilic and non-nucleophi1ic the largest paramagnetic chemical shift solvents. (A) in tetrahydrofuran (0.13 differences are observed for the methy- M); (B) in carbon tetrachloride (0.06 lene protons bound to the oxygen of the M) ; (C) , in n-octane-2Hia (O.Oit M) . The phytyl moiety and the diamagnetic monomer spectrum assignments are shown shifts observed for the a- and in A. Spectrum B is the spectrum of (3-methine protons. The methyl protons (Chi a)2. at positions 3a and 5a both show diamagnetic shifts while little or no change is observed for the protons at positions la, 10b, 10, or 0. The trinitrobenzene protons experience a large (Figure 3). In the polar solvent tet- upfield shift. These observations are rahydrofuran, Chi a occurs as the mono- consistent with the formation of a Chi solvate, Chi a»THF, but in CCU or <|-tr i ni trobenzene complex in which the n-octane, Chi a occurs as a dimer or an trinitrobenzene lies on the surface of oligomer, respectively. Evidently, in the macrocycle with two of its nitro these solvents a mobile equilibrium groups extended over the a- and nChl a * (Chi a)n exists. The extent

Vol. 5, No. 1/2 25 of aggregation is then determined by solvent, chlorophyll concentration, and temperature. The equilibrium constant for aggregation is very large, probably greater than 10' mol"1 1 for the dimer, so that the concentration of monomer Chi a in systems free of extraneous nucleophiles is very small. In polarizable, non-nucleophi1ic solvents such as carbon tetrachloride, chloroform, or benzene, Chi a occurs predominantly as the dimer, whereas in difficultly polarizable , non-nucleophi1ic solvents such as aliphatic hydorcarbons, oligomers, (Chi a)n, with n > 20 occur in 0.1 M Chi a solutions. It is, therefore, not surprising that the XH NMR spectra reflect the differences in the Chi a species present in the different solvents. lH NMR spectra of even the dimer are distorted, and for larger oligomers are obliterated. However, JH NMR studies on dimeri2ed solutions yield important information that bears on the structure of the dimer. A detailed review of chlorophyll-chlorophyll interactions can be found in ref- 1.0 2.0 3.0 4.0 5.0 6.0 8.0 9.0 10.0 MOLE RATIO C5D5N/BACTERI0CHL0R0PHYU. erence ll». Addition of a nucleophile to a solu- X tion of (Chi a) 2 in CC14 changes the H NMR spectra to an extent determined by Figure b. Titration of bacteriochloro- the molar ratios of Chi a/nucleophile phyll a (0.03 M) in benzene solution (11). When a molar excess of nucleo- with pyridine-2Hs. Chemical shifts in X phile has been added, the H NMR spec- 5, ppm relative to internal hexamethyl- trum in a non-nucleophi1ic solvent disiloxane. See Figure 1 for proton becomes identical to that of Chi a in a number i ng. neat base. Because the chemical shifts are fully assigned in the monomer spectrum, it is possible to ascertain the positions of the corresponding proton resonances in the self-aggregated (Chi acetyl function at position 2 likewise a) 2 by a titration procedure in which X experience a large downfield shift as H chemical shifts are recorded as a base is.added. All of the changes in function of Chi a/nucleophile ratio. chemical shift on disaggregation are to Such an experiment makes possible lower field, indicating that in the structural conclusions about the nature aggregate the resonances of many of the of the chlorophyll aggregate. protons are at higher fields than in A typical titration experiment, in the monomer. A reasonable hypothesis this case on aggregated Bchl a, is for this effect is a diamagnetic ring shown in Figure *• (80) . The addition of current effect on the protons experi- incremental amounts of the strong base encing a high field shift in the aggre- pyridine to a solution of (Bchj_ a) n gate. The protons fall roughly into results in a larger change for the two classes, one in which the proton a-methine resonance than for the 3 and resonances are essentially the same in o protons. The C-10 proton, the protons both monomer and aggregate, and the of the methyl groups at positions la, other in which the protons are shifted 5a, and 10b, and the CHa-group of the to varying degrees upfield. It follows

26 Bulletin of Magnetic Resonance that the chlorophyll macrocycles are the maximum differences in chemical only partially eclipsed in the aggre- shift between the aggregate and the gates. The downfield shifts observed in monomer for given protons are superim- the titration experiment can, there- posed on a structural formula. For Bchl fore, be interpreted to indicated that a aggregates, two regions of overlap the protons experiencing the largest are evident, one in the vicinity of the downfield shift are the ones most keto C=0 function, the other near the strongly shielded. The titration acetyl C=0 group. Both of these groups results can be presented in the form of must be acting as nucleophiles to the an aggregation map (Figure 5), in which Mg atom(s) of other chlorophyll molecules. The presence of two donor functions in Bchl a considerably compli- cates matters, and whether there are two populations of dinners in this system is still uncertain. 0.49 Although there is convincing evidence from 1R spectroscopy that it is the keto C=0 group in ring V that is the principal donor in the coordination interaction between Chi a molecules (12,81), Fong and Koester (82,83) proposed a symmetrical (parallel) structure for (Chi a)2 in which the carbomethoxy C=0 functions are used as donors to Mg. Additional NMR evidence on the relative donor strengths of the oxygen functions in Chi a is now available from a detailed comparison of the 0.33" aggregation behavior of Chi a and Pyrochl a (a Chi a derivative lacking a carbomethoxy group at C-10, see Figure 13 C02CH3 0 1), from a comparison of the C chemi- a62 cal shifts in Chi £«Li and (Chi a)2, c62(c20H39) and from an examination of the aggregation behavior of desoxomesochIorophy11 a (a Chi a derivative in which the keto C=0 function at position 9 is replaced by 2H) . Figure 5- Aggregation map of bacter- A titration experiment on a chloro- iochlorophy11 a from chemical shift phyll dimer or oligomer relates the differences between aggregated and chemical shifts of the protons (or 13C) monomeric bacteriochlorophy11 a. The in the aggregated species to that in numbers in the figure show the maximum the monomer, where the chemical shifts differences in chemical shift (in ppm) are fully assigned. The aggregation map between monomer and aggregate for the so constructed, thus, defines the ring indicated protons as deduced from the current shifts resulting from aggregate titration data of Figure k. The arcs formation. The results of titration indicate regions of macrocycle overlap experiments with (Chi a)2 in the aggregate resulting from coordi- and (Pyrochl a)2 are shown in Figures nation interactions by both the 6 and ~J, respectively. The numbers 2-acetyl and 9~keto carbonyl functions superimposed on the monomer structure with Mg atoms in adjoining bacterioch- give the chemical shift difference lorophyll molecules. between the corresponding proton (s) in the dimer and the monomer. A positive sign indicates that the chemical shift of the particular proton in the dimer is at higher field than in the monomer. Vol. 5, No. 1/2 27 0.02

H CH2 2.05 /7a b CH2 (0aC02

CH3 7cC02Phytyl |0b 0.6l

7cC02phytyl

Figure 6. 1H NMR titration of chloro- Figure 7• *H NMR titration of pyrochlo- phyll a (0.06 M) in carbon tetrachlo- rophyll a_ (0.06 M) in carbon tetrachloride with pyridine-2Hs. Chemical shift ride with pyridine-2Hs. Chemical shift differences (<5, ppm) between dimer and (o, ppm) differences [A(<5) = monomer [A(5) = 6monomer- <5dimer] ^monomer" 5dimer^ are Positive for are positive for upfield shifts in the upfield shifts in the dimer dimer (14) .

the dimer generated by the keto In both Figures 6 and 7» all cf the C=O***Mg interactions as suggested by shifts in the dimers are to higher Shipman et al. (84) appears, however, field, suggesting that on the average, to be consistent with the experimental all of the protons in the dimer are in findings. A similar conclusion has been the shielding region of the partner reached by Georghe et al. (78) . macrocycle. A carefully constructed The largest differences in chemical model of the Fong dimer structure put shifts between dimer and monomer are in together with carbomethoxy C=0 interac- the vicinity of ring V, and the ring tions shows that for such a structure current effects in (Chi a) 2 and the la, 3a» 4a, 10b, a-methine, and (Pyrochl a_) 2 are qualitatively similar. vinyl protons are all in the deshield- As Pyrochl a has no carbomethoxy group, ing zone of the adjacent macrocycle, it is difficult to avoid the conclusion which is not what is observed. The that it is the keto C=0 group that is results from the titration experiments, the principal donor function in both therefore, provide no support for a cases. The incremental shift in the dimer with a parallel structure cross- C-10 protons of (Pyrochl a) 2 is sub- linked by carbomethoxy C=0 interac- stantially larger than in (Chi a) 2, tions. A perpendicular structure for which indicates that ring V in (Pyrochl

28 Bulletin of Magnetic Resonance a) 2 comes closer to the center of the strength is greatest in (Pyrochl a)2, binding macrocycle than it does in (Chi which lacks a carbomethoxy group, and a)2. Indeed, from the incremental chem- weakest in 9~desoxomesoch1orophy11 a, ical shifts, (Chi a)2 appears to be which has a carbomethoxy group but no less stable than (Pyrochl a)2, and the keto C=0 function. steric hinderance from the 13C NMR has provided direct evidence 10-carbomethoxy group is a destabiliz- for the participation of the ring V ing influence rather than the driving keto function in dimerization (54,55)• force in dimerization. Figure 9 shows the results of a 13C NMR A Chi a derivative in which the car- titration experiment on (Chi a)2. The 13 bomethoxyy group is present but the incremental C chemical shifts are 9-keto group is reduced to -CH2 has again a function of ring current been synthesized by Scheer (85). In effects in the dimer, but superimposed this synthesis, the vinyl group and the is a deshielding effect expected for double bond in the phytyl chain are the carbon atom of any carbonyl func- both hydrogenated, but this is not tion participating in a coordination expected to have any significant effect interaction. By far the largest desh- on the donor properties of the mol- ielding is observed for the carbon atom ecule. As judged from the ring-current in the keto function. The carbonyl induced chemical shift of the C-10 pro- carbon in the carbomethoxy group experiences an upfield shift in the dimer. This upfield shift in the carbomethoxy carbonyl carbon effectively excludes

i i i I | i i i i • T I ' 7 this function from consideration as a donor group. The downfield shift in the carbonyl carbon atom of the propionic 6 acid side chain is also observed in all i(ppm) of its immediate neighbors, and this is consistent with a ring current rather than a coordination origin. Although X~ V . deshielded, the propionic ester C=0 resonance is sharp, suggesting that it

4 enjoys freedom of motion and, thus, does not participate to a significant extent in dimer formation.

1 1 1 I ) T 1 1 i 1 All of the available evidence thus 5 10 Molar Rotio MeOH : Chlorophyll focusses on the keto group as the principal donor function in Chi a. A similar conclusion has been reached by Ras- quain et al. (86) from an lH NMR study Figure 8. XH NMR titration experiment of the aggregation of protochlorophyl1 2 that compares aggregation in (A), H a in non-nucleophi1ic solvents. The Chlorophyll a (0.15 M) ; (0), desoxome- relative donor strengths in Chi b , and sochlorophyll a (0.59 M); and (x) , 2 Bchls a, b, £, d, and e, all of which H-pyrochlorophyl 1 a (0.0*»9 M) . Only have donor functions in addition to the the chemi.ca! shift (<5, ppm) of the C-10 9-keto group still remain to be estab- protons is shown. Both deuterio-chlo- 13 l lished. As in the case of Chi a, C rophylls had H at position C-10. The NMR is expected to be the method of 9-desoxomesoch1orophy11 a, which lacks choice. a 9-keto function, requires the small- In a system at room temperature con- est ratio of nucleophile/chlorophy11 taining both monomer and dimer Chi a, for complete disaggregation. only one set of resonance lines can be observed, which implies an averaging process rapid on the XH NMR time scale. For Pyrochl a, the XH NMR spectrum shows comparatively sharp resonances at tons (Figure 8) the aggregation room temperature and above. Decreasing Vol. 5, No. 1/2 29 electrophile, seriously perturbs the (Chi a_) 2 equilibria and the dimer present in the system experiences changes in conformation (87).

H +1.04 \2b ,H 5. Photoreaction Center Models + 0.26 There is considerable evidence that the primary electron donor in the photoreaction center of both green plants and photosynthetic bacteria is a special pair of chlorophyll molecules (88). Possible structures for the chlo- H CH2 +1.6. /+0.96H» rophyll special pair have been sug- +0.45ST2 io°ce* gested (89-93) and, despite the many \ -0.98 \ unresolved questions remaining about c CH nlc 7c °2P>>y IOb 3 -0.16 +040 the structure of the _i_n vivo photoreac- P-l t P-3 "0.49 tion center, chlorophyll special pair phy = 1J> -0.48 models have been synthesized from Pyrochl a (91), Chi a (94), and Bchl a (95)• These particular models consist of two macrocycles covalently linked through their propionic acid side chains. In the absence of extraneous nucleophiles, the covalently linked Figure 9. 13C NMR titration of chloro- pairs occur in an open configuration, phyll a dimer. The incremental chemical but in the presence of nucleophiles such as water or ethanol, which are shifts [A(o) = 5monomer- 5dimer] (<5, ppm) are shown for various carbon capable of coordination to Mg and atoms in the molecule. A negative sign simultaneous hydrogen-bonding to ring V indicates the 13C chemical shift of the keto function, folding of the linked indicated carbon atom is at lower field pair occurs to form species that mimic i n the dimer (55)• the salient features of j_n vivo photoreaction centers. 1H NMR has been used to demonstrate that in the folded configuration the linked pair has the structure suggested in references 91 temperatures lead first to increased and 92. Wasielewski et al. (95) have line broadening, but below -35° C the shown that in the folded covalently lines gradually sharpen and split into linked Bchl a special pair model, the a multitude of resonances. Such behav- 5a-CH3 and the C-10 proton experience ior is typical for an exchange process, substantial upfield shifts, while the and the line broadening at room temper- kb methyl group is shifted downfield. ature and somewhat below shows that This behavior closely parallels that (Pyrochl a)2 is close to coalescence in observed by Boxer and Closs (91) in the this temperature range. The room temp- Pyrochl a linked pair. As the chemical erature 1H NMR spectrum of (Pyrochl a)2 shift in the methyl group of the acetyl represents an average conformation to function in Bchl a_ is not perturbed by which at least two but probably several folding, it appears that only the keto more conformations contribute. An group is involved in the folding opera- attempt to define more precisely the tion (Figure 10). All of the *H NMR structures of the conformers of (Chi evidence, thus, supports the structures a_) 2 by a lanthanide-i nduced chemical proposed by Boxer and Closs (91) and shift study of Chi a in CCI4 has met Shipman et al. (92) . with indifferent success because the Recently, Boxer and Bucks (96) have shift reagent, itself a good linked through a covalent bond a

30 Bulletin of Magnetic Resonance 6. Biosynthesis Studies

In spite of the current interest in chlorin biosynthesis, NMR methods have 1.54 not been widely used. Presumably, this IIA7) reflects the lower sensitivity of NMR CH, 3.95 methods as compared to 14C tracer techniques. NMR, however, is capable of providing biosynthesis information difficult to obtain by conventional tracer 0.62 1 11.20] procedures. Applications of H NMR tc Chi a_ and Bchl a_ biosynthesis have been described by Katz and Crespi (97)- All higher green plants contain Chi a and b (Figure 1). Although it is generally accepted that Chi a is the precursor of Chi b j_n vivo , the evidence from tracer experiments on this important point is not as firm as might 1.8-2.4 3.40 (3.02) be desired. When green algae are grown (1.8-2.4) in an 1H20-2H20 mixture, both hydrogen 2 2 isotopes are incorporated. The XH/2H 4.00 ratio at different sites in the chloro- (4.3 I) phylls extracted from such organisms can readily be obtained by integration of the 1H NMR spectra of the respective methyl pheophorbides. Except for the 1H/2H ratio in the formyl group of Chi Figure 10. 1H NMR titration experiment b, and CH3 group in Chi a in position 3 that compares the chemical shifts (5, all other isotope ratios are identical ppm indicated in parentheses) in the in the two chlorophylls (Figure 11). open (in 10% pyr idine-2rls in benzene- The identical isotopic composition of 2He solution) configurations of the Chi a and b_ at all corresponding sites photoreaction center bacteriochloro- provides strong evidence for the forma- phyllide a_ special pair model tion of Chi b from Chi a, as it would bis (bacteriochlorophyl1ide a) ethylene be surprising for two independent bios- glycol diester. The 5a and 10b methyl ynthetic pathways to have identical groups are strongly shielded in the isotope effects at all corresponding folded configuration (94). sites in the two molecules. There appears to be no significant branching in the biogenetic pathways prior to the oxidation of the methyl group at postion 3 in Chi a to the -CHO group in pyropheophytin macrocycle to a pair of Chi b. The fractionation factors also covaiently linked Pyrochl a molecules. indicate some possible complications in The chemical shift changes observed on the biosynthetic pathway. If the iso- folding are consistent with a symmetri- topic composition of the vinyl group is cal, rapidly-averaging, folded configu- compared to that of the ethyl group at ration in which the pyropheophytin ring position h, no mechanism readily sug- U positioned over the 5a-CH3 and gests itself whereby an ethyl group of 3-proton of the metal-containing folded the observed isotopic composition could 1 13 be generated from a vinyl group of the pair. H and C NMR can be expected to i play an increasingly important role in observed composition even if pure H defining the structures and properties were used as the reducing agent. Either of these and other model systems now the vinyl groups at positions 2 and h under development. are produced by different reaction mechanisms, or protoporphyrin IX is not

Vol. 5, No. 1/2 31 preferentially localized in 8 sites in the phytol backbone. These results dem-

1.9 ±0.3/ V H (CH0 0.4±0.2) onstrate specific incorporation of y 13 H "V IO±Q2 CH3 1.7±0.3 [l- C]acetate, and the results are consistent with the normal terpenoid pathway to geranyIgeranyl pyrophosp- 0.5 2.3 + hate, the precursor of phytol.

I I CORRINS

CH3 1.5 ± 0.2 1.7±0.4 H3C~- H A. Structural Features 4.0 ± A corrin is a partially reduced tet- u TC02CH3 rapyrrole macrocycle in which two of hx + m 2.8±0.5 the pyrole rings are bonded directly to C02CH3 each other via their a-carbon atoms. INT. Most research involving the corrins has been performed with vitamin B12 and its biologically active analogues, and has had the goal of determining how the corrin ring is biosynthesized and how the chemical and physical attributes of Figure 11. Isotope discrimination fac- the corrin ring yield the biological tors (KIH/K2|-|) as determined by functions of B12. For this reason, we integration of the 1H NMR spectrum of shall stress the NMR spectroscopy of methyl pheophorbides prepared from Chi this special class of corrins and shall a_ and b extracted from green algae mention the literature relating to the grown in a 1:1 mixture (v/v) of synthetic corrin complexes only when it and 2H 0 (97)• applies to their biologically important 2 counterparts. Vitamin B12 (Figure 12) has three distinct components. These are the corrin ring, which binds a Co (I I I) ion via the only intermediate. The isotopic its four pyrrole nitrogens, a side composition of the methyl groups at chain that is attached to carbon-17 of positions 3 and 8a as compared to the the corrin ring and whose la and 3a methyl groups differ to an 5,6-dimethyIbenzimidazole moiety binds extent greater than the experimental an axial coordination position of the error, suggesting the possibility that Co (III) ion, and a cyanide ion bound to the methyl groups at positions la and the remaining coordination position of 5a had a different chemical history the Co(UI) ion. The cyanide is present from that of the 3a and 8a methyl in the vitamin for the technical reason groups, or that all of the methyl goups that it binds to the Co (III) ion more were not formed at the same time. Simi- tightly than most other ligands and , lar NMR experiments with Bchl a in thus, permits isolation of the molecule media of mixed isotopic composition as a single species with cyanide as the indicate similar problems requiring only ligand in this coordination posi- resolution in the biosynthetic pathways tion. This substance is properly named to Bchl a (98,99). cyanocobalamin. The coenzyme form of Ahrens et al. (100) have used 13C this vitamin in the body is NMR to study the biosynthesis of phy- 5'-desoxyadenosylcobalamin, with a tol. 13C-enriched acetate was fed to 5'-deoxyadenosy1 group substituted for Euglena graci 1 i s, the chlorophyll was the cyanide. The biological activity of extracted, and the esterifying alcohols cobalamin arises from its ability to obtained by hydrolysis and thin-layer produce a cobalt-carbon bond with a chromatography. 13C incorporation from variety of ligands at the coordination the acetate precursor was position filled by cyanide in Figure

32 Bulletin of Magnetic Resonance The 75 to 100 protons of cyanocobalamin and its various analogues give rise to a relatively large number of well resolved resonances (101-113)' The HJNOCHJC, very early experiments were severely * limited by the sensitivity and resolution of early spectrometers, and the rather low solubility of the cobalamins and cobinamides in available solvents such as trifluoroacetic acid and deuterated dimethylsulfoxide. By comparing HjNOCHjC-t'8 the spectrum of cyanocobalamin with those of its various isolated substit- OC-CHJCH^CHJ,' CH3 uents and of synthetic corrin com- NH .'* plexes, it was possible, however, to

CH2 assign several chemical shift ranges to specific functional groups. The early observations yielded a number of interesting results: i) the chemical shifts of the protons of the 5,6-dimethylbenzimidazole moiety depend on whether this ring is protonated and/or coordinated to the Co (I I I) ion; HO-CH ii) the chemical shift of the proton on 2 the methine bridge C-10 depends on the ligands attached to the cobalt; iii) the proton on C-10 exchanges rapidly in Figure 12. Structure of vitamin B12. trifluoroacetic acid; and iv) the cor- The coordination position of the CN rin system does not exhibit an aromatic above the corrin ring is referred to as ring current. the fifth coordination position and The advent of superconducting NMR that to which the benzimidazole nitro- spectrometers made it possible to gen atom is coordinated below the cor- record useful spectra on dilute solu- rin ring is the sixth coordination tions in 2H20, and this made possible position, in agreement with the nomenc- the assignment of the entire *H NMR lature of Brodie and Poe (I 12). The spectrum and the characterization of ligands in these coordination positions the parameters that regulate the bind- also are referred to in the text as the ing of ligands to the fifth and sixth axial ligands. The indicated carbon coordination positions of the Co(lll) atoms of vitamin B12 are derived from ion (Figure 12) (108-113). The proton carbon atom-5 of 6-aminolevulinic acid on C-10 of both cobalamins and cobinam- (0) and from the methyl group of methi- ides, which readily undergoes electro- oni ne (*) . phi lie substitution reactions (HA), was demonstrated to exchange with 2H20 under acid conditions at a rate that is intermediate between those of porphyrins and chlorins (107). ln addition, 12. When the 5»6~dimethyIbenzimidazole- it was demonstrated that the chemical nucleotide is removed so that both shifts of the protons on the axial axial coordination position of the ligand in the sixth coordination posi- Co(lll) ion are available for binding tion of the cobalt and on C-10 of the exogenous ligands, the complex is corrin ring depend on the identity of called a cobinamide. the ligand in the fifth coordination position. The length of the Co-N bond B. H\ NMR Chemical Shift Assignments between the Co (II!) ion and the dimeth- and Appli cations yIbenzimidazole appears to be a

Vol. 5, No. 1/2 33 function of the nature of the other associated with a paramagnetic interme- axial ligand, as does the electron den- diate when the benzimidazole is not sity on the cobalt. The electron den- coordinated (113)- Methyl cobinamide sity of the cobalt appears to be delo- appears to bind a molecule of water to calized partially to the C-10 position its sixth coordination position, of the corrin ring, and indeed there is whereas 5'-deoxyadenosy1 cobalamide may a correlation between the chemical not (112,113)• Aquocyanocorrins can shift of the C-10 hydrogen and the exist in two isomeric forms in which energy of the first electronic absorp- the water and cyanide occupy the fifth tion band (107,112). and sixth coordination positions in Since the protons on the 5'"carbon either order (123,124). The only of the 5'"deoxyadenosy1 moiety of the effect on the XH NMR spectrum of these coenzyme appeared to be involved in two isomers appears to be a small hydrogen transfer reaction and reduc- splitting of the resonance of the pro- tion of the Co (III) ion of cobalamin ton on C-10. appeared to occur in both hydrogen and The resonance in the XH NMR spectra methyl transfer reactions (115~H8), of the cobalamins at ~0.5 ppm from TMS the XH NMR spectra of these complexes was assigned to the methyl group on C-l were studied. The S^protons of the of the corrin ring, and the shielding 5'-deoxyadenosy1 ligand in the coenzyme was shown to arise from the aromatic were found not to exchange readily with 2 ring current of the 5.6-dimethylbenz- H20 and to be magnetically nonequiva- imidazole moiety coordinated to the lent (108,109,113), but the results of Co (I I I) ion (Figure 13) (107). Methyl a more recent experiment with x3 groups of alkyl ligands bonded to the [5'" C]-5'-deoxyadenosy1coba1 am i n cobalt were demonstrated to resonate at raise the possibility that the earlier even higher field than the C-l methyl assignments for the 5'protons were group (112,125)• Comparison of a large incorrect (119)- In any case, nonequi- number of cobalamins and cobinamides valence of the protons of the cobalt- (112) permitted almost complete assign- bound methylene group has been observed ment of the proton resonances of the with alkyl ligands other than methyl groups on the corrin ring (Table 51-deoxyadenosine in the fifth coordi- 13) • nation position (112). The cobalamins with cobalt in the Co (I) and Co (I II) C. j_?_C NMR Chemical Shifts and Applica- oxidation states were found to be diam- tions agnetic, but the *H NMR spectra were found to be paramagnetically broadened The application of 13C NMR spectros- when cobalt was in the Co (I I) oxidation copy to the investigation of B12 and state (112). Spectral line broadening its analogues with natural-abundance appears to be a consequence of the rel- 13C did not become feasible until the atively long relaxation time of the late 196O's. The detection and assign- unpaired electron of the Co (II) ion ment of the natural-abundance X3C reso- (120-122). Furthermore, it was found nances required large diameter sample that the 5,6-dimethylbenzimidazole tubes and pulse-Fourier transform tech- group does not bind to the sixth coor- niques to increase sensitivity and dination position of the Co (I) cobala- facilitate measurements of spin-lattice min although it does so in the Co (III) relaxation times (Ti). From chemical cobalamins. The 5.6-dimethylbenz- shift comparisons, Ti values, 13C-31P imidazole group of Co (III) coenzyme B12 spin-spin coupling constants, and the is in a dynamic equilibrium with the results of off-resonance single-fre- coordination and uncoordinated states; quency proton decoupling experiments the first-order rate constant for the Doddrell and Allerhand were able to breaking of the Co (I I I)- benzimidazole provide surprisingly complete assign- coordination was estimated as somewhat 1J x ments for the C NMR spectra of cyano- larger than 550 s (113). This equi- coba1 am i n, 5'-deoxyadenosy1coba1 am i n librium is dependent upon pH and temp- and a number of their analogues erature, and does not appear to be (126,127). The.region of the spectrum

Bulletin of Magnetic Resonance of dicyanocobalam in were assigned to the broad resonance centered at 55 ppm. Broadening was suggested to be due to scalar coupling to the 5'Co nucleus, which has a quadrupole moment. Unlike the XH NMR spectra of the two isomers of aquocyanocorrins, the 13C chemical shifts of aquocyanocobyric acid were found to be very sensitive to the isomeric orientations in which the water and cyanide can bind to the cobalt. The observed differences in chemical shifts between the two isomers were Hf attributed to both electronic differences and conformational changes of the corr in r i ng. Historically, the use of specific 13C labelling and 13C NMR for detection of the label that followed the original assignments of Doddrel1 and Allerhand fell into two categories. One of these was concerned with establishing the biosynthetic pathway to Bt2, and the other consisted of studies on the effects of axial ligands on the biological activity of B12.

I. 13C NMR of the Corrin Ring Figure 13- The orientations of the methyl group of vitamin B12 with The early steps in the biosynthetic respect to the corrin and pathway to the corrins and to the por- 5,6-dimethylbenzimidazole rings. The phyrins had been shown to be quite sim- pro-R methyl group at carbon atom 12 ilar (128-130). It was known that the lies below the plane of the corrin ring 5-carbon atom in 6-aminolevulinic acid and the pro-S methyl group lies above. (Figure ]k) was the precursor of at The acetamide and propionamide side least seven carbon atoms in the corrin chains have been omitted for the sake ring of 612, and of eight carbon atoms of clar i ty. in the porphyrins. The mechanism by which pyrrole ring D is incorporated into the corrins and prophyrins was the subject of much speculation, and the origin of the methyl group on C-l of from about 85 to 100 ppm upfield from the corrin was uncertain. It was known CS2 was assigned to the methine carbons that one of the two methyl groups on that bridge pyrrole rings A, B, and C. C-12 of the corrin ring arises from C-10 was assigned to the resonance at decarboxylation of an acetic acid side 100 ppm, whereas C-5 and C-15 were chain and that at least six of the assigned to the resonances at 86 to 89 seven remaining methyl groups on the ppm because they are bonded to methyl corrin ring arise from the methyl group groups. The unsaturated carbon atoms of of methionine, but the general concen- the corrin ring directly bonded to the sus was that the methyl group on C-l pyrrole nitrogen atoms were assigned to should arise from the 5~carbon atom of a downfield spectral range that over- o-aminolevulinic acid, as does the laps that of the amide carbonyl region, 5-bridge carbon atom of the porphyrins and the methyl groups were assigned to (Figure 14). For technical reasons, this latter point could not be proven the spectral lines above 170 ppm rela- 14 tive to CS2. The cyanide carbon atoms with C-labei1 ing experiments (130). Vol. 5, No. 1/2 35 PPM(CS2) 100 200 P0RPHYR1NS I . l 200 100 0 PPM(TMS)

Figure 14. The labelling pattern that Figure 15- The proton-decoupled 13C NMR is expected when the ring structures of spectra of vitamin B12 in water. The porphyrins and corrins are biosynthes- spectrum of (A) vitamin B12 synthesized i2ed from 6-aminolevulinic acid that is from o-amino[5-13C]levulinic acid; and labelled isotopicaily in carbon atom 5 (B) vitamin B12 synthesized from (0). The curving dashed arrow repre- L-[methyl-13C]methionine. CS2 is the sents the early hypothesis that the external reference, but the chemical corrin ring was biosynthesized directly shift is presented in ppm relative to from a linear tetrapyrrole intermedi- both CS2 (top scale) and tetramethylsi- ate. It has since been demonstrated lane (bottom scale) to permit easier that uroporphyrinogen III is an intercomparison of spectra in the references mediate in the biosynthetic pathway to (see text) . the biological corrins (48-50).

carbon atoms 5 and 15 of cyanocobalamin To overcome this complication, we added and demonstrated that carbon atom 5 o-amino[5~13C] levulinic acid to a cul- resonates at lower field. The complex ture of Propionibacter ium shermani i, 13C-13C spin-spin spli tting pattern isolated the B12 produced as cyanocoba- observed for C-15 provided direct evi- lamin, and recorded its 13C NMR spec- dence that the pyrrole ring is turned trum (Figure 15A) (131) • This spectrum over during its incorporation into the confirmed the labelling pattern discov- corrin ring, as also occurs in porphy- ered by radiotracer techniques and con- rin biosynthesis. This experiment firmed the assignments for these carbon yielded the unexpected result that the atoms by Doddretl and Allerhand (126). methyl group on C-l of the corrin ring From the 13C-13C spin-spin splitting did not contain an appreciable amount 13 patterns, we distinguished between of C-label. Similar results were

36 Bulletin of Magnetic Resonance obtained by Scott et al. (132), who also assigned the 13C resonances of the methylene carbon atoms in the acetamide and propionamide side chains of the corrin ring of cyanocobalamin and the methyl group on C-12 of the corrin ring that arises from decarboxylation of an acetic acid side chain (132,133). It is interesting to note that this methyl group on C-12, rather than the methyl group on C-l as assumed by Doddrell and Allerhand, resonates at abnormally low- field compared to the other methyl carbon atoms in cyanocobalamin. This has been ascribed to the lack of a y effect for this one methyl group (see below) (133). Specific 13C chemical shift assignments were reported recently for the carbon atoms of several derivatives of cyanocobalami n (13M. Realization that the methyl group on C-l of the corrin ring does not arise from the 5~carbon atom of 6-aminolevulinic acid led to the immediate assumption that seven, rather than six, methyl groups arise from the methyl group of methionine. When Scott et al. (132,133) added L-[methyl- 13C]methionine to a culture of Propion- Figure 16. The proton-decoupled l3C NMR ibacter ium Shermani i and isolated the spectra of vitamin B12 synthesized from B-. 2 produced as cyanocobalami n, the L-[methyl-13C]methionine. The spectra methyl region of the 13C NMR spectrum were obtained in water (A), in 0.1 M was found to exhibit only six enhanced KCN (B), and in aqueous HC1 (pH > 1) 13C resonances. However, addition of (C) . The chemical shift values are excess cyanide to the sample to produce given in ppm downfield from tetrame- dicyanocobalamin resulted in the thylsi 1ane. appearance of a seventh peak in the upfield 13C NMR spectrum, and this was taken as proof that the methyl group on C-l of the corrin ring arises from the methyl group of methionine. Brown et This latter experiment pointed out al. (135) reproduced this work, and one of the apparent anomalies in the XH also demonstrated that protonation of NMR spectrum of cobalamins. The labora- the benzimidazole ring and substitution tories of both Scott and Battersby dem- of a water molecL'le at the sixth coor- onstrated that the methyl group of dination position of the cobalt also methionine gives rise to the pro-R produces seven enhanced resonances in methyl group on C-12 of the corrin ring the upfield region of the 13C NMR spec- (i.e., the methyl group cis to the pro- trum (Figures 15B and 16) . In addition, pionamide side -chain on C-13» marked single-frequency proton-decoupled 13C by an asterisk in Figure 12). Scott et NMR spectra were recorded for all three al. (133) produced dicyanocobinamide of these cobalamin complexes, which and dicyanoneocobinamide, in which the permitted assignment of all seven configuration of C-l3* in ring C is enhanced resonances on the basis of the reversed (136), from cyanocobalaminv existing assignment of the methyl that had been biosynthesized with groups in the XH NMR spectrum of B12 L-[methyl-i3C]methionine. The l3C reso- (Figure 17 and Table 13). nance of the enriched methyl group of Vol. 5, No. 1/2 37 the dicyanocobinamide was found to resonate at higher field than that of the dicyandneocobinamide. This shielding was attributed to the steric interactions between the labelled methyl group and the adjacent propionamide side chains, i.e., the y effect (137,138). Battersby et al. (139,UO) performed the same label 1 ing experiment but isolated the imide of ring C by ozonolysis and compared the XH and 13C NMR spectra of this isolated ring with those of a chemically synthesized standard. By combining the assignment of Scott and Battersby for the 13C resonances of the methyl groups on C-12 with the single-frequency proton-decoupling experiments, it is possible to assign the proton resonances of these two methyl groups, as shown in Table 13. The assignment of the peak at 120 ppm for the protons of the pro-S C-12 methyl group, which lies above the corrin ring in Figure 13, agrees with the observation that this peak does not PPM change position when cyanide is substi- N N N — tuted for the benzimidazole ring in the sixth coordination position of the cobalt below the corrin ring. However, this assignment leads to the unexpected Figure 17. The *H NMR spectra of vita- conclusion that of the two methyl min B12. A spectrum of vitamin B12 with natural abundance 13C; vitamin B12 syn- groups on C-12 of cyanocobalamin, the 13 one lying above the aromatic benzimida- thesized from L-[methyl- C]methionine. zole ring is the less shielded. In the The chemical shift values are given in XH NMR spectrum of dicyanocobalamin, 6, ppm downfield from tetramethyIsi- the C-l methyl group is deshielded by lane. The two arrows about the resonance 0.46 ppm indicate the satellite displacement of the benzimidazole from l3 1 the cobalt, and one other methyl group peaks due to C- H spin-spin split- (which is the C-2, C~7, pro-R C-12, or ting. C-17 methyl group) is more shielded, so that its chemical shift is the same as that of the pro-S C-12 methyl protons (Table 13). These apparent anomalies may be explained by anisotropic shield- uroporphyrinogen III and sirohydochlo- ing properties of the bound cyanide, rin are intermediates in the biosyn- but this has yet to be demonstrated thetic pathway to corrins (144-147) . fully. Since this work has been the subject of recent reviews (148-150), it is not The subsequent 13C NMR experiments further discussed here. involving the corrin ring have utilized the spectral assignments presented 2. 13C NMR of the Axial Ligands above for studying the incorporation of various 13C-labelled intermediates into 13C NMR spectroscopy has been used 13 13 the corrin ring. This work has led to to investigatg e - CN,, - CH3 (125)(5 , the discovery that the methyl groups of 13 13 13 - CH2- COOH (151) and [5'- C]-5'"de- the corrin ring that arise from methio- oxyadenosine (119,152) bonded to the nine are incorporated without proton fifth and sixth coordination positions exchange (l4l-li»3), and that both of coba cobinamide, and 38 Bulletin of Magnetic Resonance o

•z. Table 13 o 'H NMR and 1]C NMR Chemical Shifts of Vitamin B i ? and 'H Chemical Shifts at which Nuclear Overhauser Enhancement (NQE) 1s maximum for '3C Resonances.

Chemical Shifts in 5, ppm from TMS in H?0, O.1M KCN and HCL aq (pH<1)

KCN HC1

1 2 3 4 5 6 7 8 9

Position of Methyl Group 1n Assigned Vitamin Bi? (see Figs. 12,13) 'H NMR 1>C NMR NOE 'H NMR ' 'C NMR NOE 'H NMR 1'C NMR NOE

C-1 0.46b 20.059 0.46 1 .44 22.53 1 .44 1.81 24.589 1 .8 C-12 (Methyl group trans'" to 1 .20 1 .26 1.15 propionamide on C-13 Methyl of aminopropanol 1.26b 1.31 1 .20 C-2d , C-7d, C-12, (c1s«= to 1 .39 17.689 1 .39 1 .22® 18. 119 1 .24 1 .31 19.4O9 propionamide on C-13), and 1 .41 16.719 1 .41 1 .37 17.259 1.41 1 .54 18. 339 (1.3-1.6)' C-17d 1 .45 20.05 1 .48 1 .46 19.40 1 .46 1 .59 20. 16 1 .87 19.84 1 .88 1 .74 19.73 1 .74 1.81 21 .56 1.8e Methyl groups of benzimidazole 2.26b 2.36 2.35 2.38 C-5 and C-15 2.54b 16.28h 2.54 2.26 15.63 2.23 2.38 16.06 2 .40 2.58b 15.95h 2.60 2.30 16.06 2.59 2.40 16 .60 2.44

a) The nuclear Overhauser enhancement data were obtained by single-frequency experiments by Brown et al. (135). b~5 Chemi-" cal shifts assigned by Hill et al. (106,107). c) ']C chemical shifts assigned by Scott et al . (133) and Battersby et al . (139). d) Individual assignments cannot be made among the 'H NMR peaks of these methyl groups. However, the ' 'C-resonances are related to the 'H NMR assignments as determined by NOE. >H chemical shifts assigned by Brodie et al. (112). e) We cannot differentiate the assignments of these two proton resonances. f) Three chemical shifts were found 1n the region of 1.3-1.6 ppm. However, no assignments of the proton resonances to the ''C resonances can be made. g) The 1]C resonances of these three methyl groups appear to shift to lower field when the dimethyl benzimidazole group 1s replaced by a less bulky CN or H?0 1igand in the sixth coordination position. h) Hogenkamp et al. (152) have assigned 516.28 to the C-5 carbon atom. epicobalamin (153)• The use of severe restriction to rotation about 13C-labelled cyanide confirmed the the carbon-cobalt bond for both these assignment of Doddrell and Allerhand coordinated ligands. The pKa of the (126) for the cyanide carbon in the carboxymethyl ligand is unusually high natural abundance spectrum of cyanoco- (pKa = 7.2), and both.the restricted balamin and provided evidence that rotation and the high pKa value have aquocyanocobinamide binds the cyanide been attributed to hydrogen bonding and water ligands in two isomeric ori- between the carboxyl group and the ace- entations, as does aquocyanocobyric tamide side chains on the periphery of acid (127). 13C-cobalamin was observed the corrin ring. The restricted motion to yield one relatively sharp enhanced of the 5'-deoxyadenosyl moiety was 13C NMR resonance in the upfield region attributed to a combination of steric of the spectrum, with a chemical shift interactions and intramolecular hydro- dependent on the pH of the solution. gen bonding. It is of interest to note This pH dependence was attributed to that the Ti values of the displacement of the 5,6-dimethy1- E13C]methylcorrinoids became shorter benzimidazole moiety by water at low when the benzimidazole moiety is pH. I3CH3-cobinamide gives rise to two removed from the coordination sphere of enhanced 13C resonances of equal inten- the cobalt. sity in the upfield region of the spec- The chemical shifts of the axial trum. These have been assigned to ligands were found to be highly depen- methyl groups bound to two different dent on the identity of the other axial isomers in the fifth and sixth coordi- ligand (the trans effect) in both the nation positions of the cobalt, and the cobalamins and cobinamides. Replace- fact that the two resonances are of ment of the dimethylbenzimidazole equal intensity has been taken to mean moiety of methylcobalamin by water at that there is no steric interference low pH causes the methyl 13C resonance against the binding of a methyl group to shift to higher field as noted to either coordination site. above. By investigating a number of In contrast to the observation of substituted methylcobinamides it was Doddrell and Allerhand (126), the 13C possible to make the more general resonances of both the -13CN and -13CH3 observation that substitution of a weak ligated to cobalamins or cobinamides ligand by a strong field ligand in the were found to be relatively narrow. sixth coordination position leads to a This was attributed to a quadrupolar substantial downfield shift and reduc- contribution that dominates the spin- tion of the 13C-1H coupling constant of lattice relaxation rate of s'Co and the methyl group in the fifth coordina- thereby obliterates the cobalt-carbon tion position. The highest chemical spin-spin splittings. Chemical exchange shift and JQ-H °f the methyl ligand of the methyl group of methylcobalamin were observed with a molecule of H2O in and methylcobinamides was found not to the sixth coordination position, and contribute to the line width of the were found to decrease in the order H2O methyl 13C resonance, but chemical > pyridine > benzimidazole > CN" > CH3 exchange between the "base on" and upon substitution of these ligands in "base off" forms of methylcobalam in at the sixth coordination position. Both 13 pH values near the pKa of the dimeth- the chemical shift of the C resonance ylbenzimidazole moiety was found to and the 13C-1H coupling constant of the have a marked effect on the line width methyl group in the fifth coordination of the methyl 13C resonance. position are linearly correlated with From the long T1 value of the methyl the energy of the (3-vibrational compo- group of methylcobaI am in it was con- nent of the first electronic absorption cluded that rotation about the carbon- band. cobalt bond is rapid and unrestricted. The chemical shift of the proton on The same is not true, however, for the C-10 of the corrin ring (the only pro- 13 13 carboxymethyl (- CH2- C00H) and tonated methine bridge) correlated with 5'-deoxyadenosyl ligands. These have the energy of the first electronic very short T1 values, which indicate absorption band as discussed above

Bulletin of Magnetic Resonance (107,112), and this was taken to sug- substitution of a small ligand for the gest that the charge density of the bulky dimethylbenzimidazole ring in the cobalt might be delocalized to this sixth coordination position of the methine bridge. Delocalization of cobalamins results in downfield shifts charge from the cobalt has been demon- of the 13C resonances of four of the strated by 13C NMR to occur with all methyl groups on the periphery of the three methine bridges (cis effect). corrin ring. The most shifted of these When a weak field ligand in the sixth resonances arises from the C-l methyl coordination position is replaced by a group (Table 13) (135) and the least strong field ligand, all of the methine shifted has been assigned to the methyl 13C resonances shift to higher field. group on the C~5 methine bridge (152). In the methylcobalamide series the The two remaining shifted resonances methine 13C resonances shift to high (Table 13) may arise from the other two field in the order H2O < pyridine < methyl groups that extend from the cor- cyanide, and similar correlations have rin ring on the same side as the been noted with cyanocorrinoids and dimethylbenzimidazole moiety (i.e., the alkylcobalamins. The chemical shifts of methyl groups on C-7 and C-12 (pro-R) the 3-vibrational component of the in Figure 13). These downfield shifts first electronic absorption band, as in resulting from substitution of a less the case of the 13C NMR resonance of bulky ligand in the sixth coordination the methyl ligand in the fifth coordi- position have been attributed to a nation position. However, introduction reduction in steric compression (the y of ligands in the sixth coordination effect) (152) . 13 position that cause the C resonance On the basis of comparisons of the of the methine bridges to shift to lH and 13C NMR spectra of the various 13 higher field cause the C resonance of cobalamins and cobinamides that have the methyl ligand in the fifth coordi- been studied to date, it appears that nation position to shift to lower there are at least three parameters field. that determine the biological activity In the epicobalamins the of coenzyme B12 (51-deoxyadenosyl- 5,6-dimethylbenzimidazole moiety is cobalamin) and methyl cobalamin. These coordinated more strongly to the are as follows: i) the charge density cobalt, the cobalt is more electronega- on the cobalt and its delocalization tive, and the carbon-cobalt bond in the over the two axial ligands and the fifth coordination position is stronger methine bridges of the corrin ring than in the cobalamins. These differ- (15^-160); ii) steric interactions ences have been attributed to both between the dimethylbenzimidazole steric and electronic changes that may nucleotide and the propionamide side result from reversing the configuration chains at C-8 and possibly C-13, and by which the propionamide side chain is the methyl groups at C-l, C-5, C-7, and bonded to C-l3 of the corrin ring, but C-12 (pro-R) that project from the the exact mechanism appears to be ster- periphery of the corrin ring on the icaily blocked since the intensities of same side as this nucleotide the two methy) resonances arising from (133,135,152,161); and, iii) hydrogen the two isomers of aquo[13C]- bonding between the axial ligand in the cyanocobinamide are in a ratio of 95^5- fifth coordination postion and the ace- In contrast, the chemical shifts of tamide side chains on this side of the the 13C nuclei on the remainder of the corrin ring. When the dimethylbenzimi- periphery of the corrin ring appear to dazole group of the cobalamins is be much less affected by the charge replaced by a weaker field ligand, the state of the cobalt, or at least other overlap between the sp3 orbitals of the variables appear to play more prominent carbon atom in the fifth coordination 2 roles. Axial ligands do affect the position and the ks and 3dz orbitals chemical shifts of the 13C atoms on the that are localized in the axial bonds periphery of the corrin ring, but the of the cobalt ion is expected to mechanism by which the effects arise decrease (155). Decreased overlap is may be different. For example, expected to increase the polarizabi1ity

Vol. 5, No. 1/2 k) of the cobalt-carbon bond, with elec- D. NMR Studies with Other Nuclei tron density shifted from the cobalt ion to the carbon atom, and thereby We are aware of only one investiga- weaken the cobalt-carbon bond. Mecha- tion of cobalamins by 15N NMR (I63) and nisms that have been suggested to cause three with 31P NMR (164-166) . The 15N the observed change in chemical shifts NMR spectrum of cyanocobalamin exhibits of the axial ligand in the fifth coor- seven resolved amide nitrogen reso- dination position to higher field nances in the region 256.8-268.2 ppm, include increased electron density on and a cyano group nitrogen resonance at the carbon atom, variations in spin 8O.9 ppm upfield from external 0.1 M pairing between p electrons, the extent 2H1SNO3 in 2H20. The four pyrrole ring of p orbital occupation, changes in nitrogen atoms were not observed, pos- radial distance of electrons from sibly because of long relaxation times, screened nuclei, and variations in small nuclear Overhauser enhancement excitation energies for mixing ground values, and coupling to cobalt. with excited state wave functions The first 31P NMR studies (164,165) (125). The decrease in electron den- are an attempt to use the 3XP atom in sity on the cobalt ion that is expected the cobalamins as a probe of events to occur upon substitution of the that occur at the corrin site of rela- dimethyibenzimidazole group by a weaker tively low molecular weight, field ligand increases the demand by Bi2-dependent enzymes such as the ribo- the cobalt ion for the electron density nucleotide reductases. The chemical that is delocalized over the methine shift of the 31P atom of cobalamins in bridge carbon atoms of the corrin ring, solution was found to depend on the which in turn decreases the charge den- coordination of the dimethyibenzimida- sity of these methine carbon atoms and zole moiety to the cobalt atom, but it shifts their 13C NMR resonances to is relatively insensitive to the iden- lower field. Steric interactions tity of the ligand in the fifth coordi- between the dimethyibenzimidazole nation position. The line width of the nucleotide and the side chains are 31P resonance increases when the cobalt expected to reduce the strength of the atom is reduced with d,1-penici1lamine. cobalt-nitrogen bond with the benzimi- It was demonstrated that 31P NMR can be dazole ring, and, by the mechanism used to investigate the pH-dependent described above, to weaken the cobalt- properties of cobalamin and the coordi- carbon bond to the ligand in the fifth nation state of cobalamins when bound coordination position. Thus, a reason to bovine serum albumin and the deter- for the methylation of seven carbon gent SDS. atoms on the corrin ring may be to pro- 31 vide sufficient destabi1ization of the P NMR spectroscopy also has been cobalt-carbon bond to the trans ligand used (166) to characterize a recently in the fifth coordination position to discovered isomeric form of vitamin make possible the biological roles of B12. The UV-visible absorption spectrum B12. This possibility is supported by of this new isomer is the same as that 1 the observation by XH and 13C NMR that of native B12, but Mossbauer and H NMR such steric factors have a rather pro- results (167) suggest that the new nounced effect on the rate of benzimi- isomer has a small out-of-plane dis- dazole dissociation in methyl cobalamin placement of the cobalt and concomit- (162). Hydrogen bonding between the tant change in the conformation of the ligand in the fifth coordination posi- corrin ring and benzimidazole group. tion of the cobalt and the acetamide Spin-lattice relaxation measurements of 31 side chains that project on this side the P nucleus in both isomeric forms of the corrin ring appear to be negli- of the paramagnetic cob (I I)alamins and gible in methylcobalamin, but can be diamagnetic cob (I I I)alamins appear to expected to play some role in determin- corroborate these conformational dif- ing the stability of the cobalt-carbon ferences (166). The spin-lattice relax- 31 bond in coenzyme B12. ation times of the P nucleus in the cob(!l)almin isomers are dominated by dipolar interaction with the

Bulletin of Magnetic Resonance paramagnetic cobalt (I I) ion. Differ- substituted for cobalt ences in the measured Ti values indi- (104,148,176,177) . Furthermore, the cate that the Co(ll)-31P distance red and yellow, metal-free corrins that becomes longer when the new isomer is were first isolated from photosynthetic formed. In the diamagnetic bacteria by Toohey (178,179) have now cob (I I I)alamins, the spin-lattice been subjected to NMR analysis relaxation time of the 31P nucleus is (180-182). The predominant metal-free determined mainly by dipolar interac- corrins that are excreted into the cul- tion with protons on the neighboring ture media by Rhodopseudomonas spher- ribose and aminopropanol groups. Varia- oides are hydogenobyrinic acid c-amide tions in Ti values of the 31P nucleus and hydrogenbyrinic acid a,c-diamide in these diamagnetic complexes support (180)• These descobaltocorrinoids are the contention that the "puckering" of formed when there is a deficiency of the corrin ring is different in the new cobalt in the culture medium and have isomer. These conformational differ- no known function. Broken cell prepa- ences appear not to alter the elec- rations of Propionibacterium shermani i tronic environment of the 31P nucleus and R. spheroides do not insert cobalt since its chemical shift in the new into hydrogenobyrinic acid a,c-diamide isomer is the same as that in native (180). However, chemical insertion of B12. cobalt into hydrogenobyrinic acid c-am- The s'Co NMR spectrum of a cobalamin ide yields, in addition to cobyrinic has been reported (165)• Furthermore, acid c-amide and 13~epi-cobyrinic acid it has been suggested that 5'Co nuclear c-amide, small amount of a blue quadrupole resonance spectroscopy could cobalt-containing corrin. This blue have great value in probing the envi- corrin has been identified to be ronment about the cobalt in vitamin 18,19-didehydrocobyrinic acid c-amide B12, methyIcobalamin, and coenzyme B12 (l8l) and may be an intermediate in the (169) • biosynthesis of vitamin B12 (180). Newer departures include characteri- E. Summary zation of the chemical reactivities of vitamin B12 and its derivatives. The The investigations of the biosyn- effects of various ligands on the pho- thetic pathway to B12 and the chemical to I ability of alkylcobinamides and and enzymatic activities of B12, which coenzyme B12 have been determined stimulated the application of NMR spec- (183-185). It has been demonstrated troscopy to the corrins, cannot be con- that mercuric ion and platinum com- sidered complete. Now that the basic plexes such as cis-diami nodiaquopiati- NMR techniques have been developed, the num(ll) displace the benzimidazole ring 13 assignments of the C resonances of from the cobalt of a IkyIcobalamins the corrin ring have been made, and the and/or coenzyme B12. Methylcobalamin chemical synthesis of potential meta- 13 has been shown to be capable of methy- bolic intermediates with specific C lating mercuric ion and PtCl6~2 labels is possible (148,167), rapid (186-188). No chemical reactivities of advances in our understanding of his various corino id complexes were com- biosynthetic pathway can be expected. pared in a recent review (189)• There continue to be reports and Another area of ongoing interest is reviews in which chemical analogues of activity of B12 as an enzyme cofactor. cobalamins are characterized with NMR The relative binding of coenzyme and spectroscopy (171-17*0 • Complete inter- 13 substrate to ethanolamine ammonia-lyase pretation of the C NMR spectrum of has been investigated (190). Nonenzy- dicyanobyrinic acid heptamethyl ester matic modelling of the coenzyme has been reported (175)- A more flexi- B12-dependent isomerization of methyl- ble form of cobalamins with a different ma I onyl coenzyme A to succinyl coenzyme conformation of the corrin has been A has been demonstrated (191-193) and discovered (167), and analogues of cobalamins have been characterized in the reversible cleavage of the cobalt- which various metal ions have been carbon bond of coenzyme B12 by methyl- ma I onyl CoA mutase has been observed

Vol. 5, No. 1/2 (191*) • Coenzyme B12 that is stereospe- 3Govindjee (Ed.), Bioenergetics of cifically deuterated in the 5'-position Photosynthesis, Academic Press, New (19*4,195) was found to lose deuterium York, 1975. to the solvent and to undergo scram- 4G. R. Seely, in The Chlorophylls, L. bling of deuterium between the two P. Vernon and G. R. Seely, Eds., Aca- diastereotopic 5'-positions in the demic Press, New York, 1966, pp. presence of the mutase (195)• The 67-109. reaction appears to involve cleavage of 5A. H. Jackson, in Chemistry and Bio- the cobalt-carbon bond and conversion chemistry of Plant Pigments, T. W. of the 5'-carbon atom into a torsiosym- Goodwin, Ed., Academic Press, New York, metric group. The exchange reaction is 1976, Vol. 1, pp. I-63. catalyzed by the methylmalonyl-CoA 'E. D. Becker and R. B. Bradley, J. mutase but occurs without the partici- Chem. Phys. 3_[, 1 i* 13 (1959) • pation of the substrate. It has been 7R. J. Abraham, Mol . Phys. 4, \k5 suggested that the mechanism of diol- (I960. dehydrase might also be investigated with the use of coenzyme B12 stereospe- . J. Abraham, K. M. Smith, D. A. Goff, cifically deuterated in the 5'-position and J.-J. Lai, J. Am. Chem. Soc. (196). These enzyme studies depend on (1982). *T. R. Sul- the recent assignment of the two Janson, A. R. Kane, J. F 1ivan, K J. 5'-protons of coenzyme B12 to XH reso- Knox, and M. E. Kenney, Am. Chem nances at 0.6 ppm and I.5 ppm Soc. 9_i, 5210 (1969) . 'R. B. 09^i195)• These assignments were made Woodward and J. Skaric, J. Am. Chem. Soc. possible by the improved resolution of 8_l, 1*676 (1961). 10J. W. the new higher-field superconducting Mathewson, W. R. Richards, NMR spectrometers. One might expect and H. Rapoport, J. Am. Chem. Soc. 85, that these investigations of enzyme 36** (1963) • X1 mechanisms will be facilitated by other G. L. Closs, J. Katz, F. C. Pen- technical advances, such as, for exam- nington, M. R. Thomas, and H. H. ple, the recent report that the XH NMR Strain, J Am. Chem. Soc. 85_, 3809 spectrum of as little as 2.5 mM vitamin (1963). B12 can be measured in 95% H2O with a 12J. J. Katz, R. C. Dougherty, and L. multipulse sequence called the 2-1-^ J. Boucher, in The Chlorophylls, L. P. pulse (197)- It seems safe to say that Vernon and G. R. Seely, Eds., Academic many more applications of NMR spectros- Press, New York, 1966, pp. 185-251. copy to the investigation of the cor- 13H. Scheer and J. J. Katz, in Por- rins can be expected. phyrins and Metalloporphyrins, K. Smith, Ed., Elsevier, Amsterdam, 1975. PP. 399-52^. 14J. J. Katz, L. L. Shipman, T. M. Cotton, and T. R. Janson, in The Pro- phyrins, D. Dolphin, Ed., Academic ACKNOWLEDGMENTS Press, New York, I978, Vol. 5C, pp.

Work performed under the auspices of 15T. R. Janson and J. J. Katz, in The the Office of Basic Energy Sciences, Porphyrins, D. Dolphin, Ed., Academic Division of Chemical Sciences, U.S. Press, New York, 1979, Vol. **B, pp. Department of Energy. 1-59- l*R. J. Abraham, A. H. Jackson, G. W. Kenner, and D. Warburton, J. Chem. REFERENCES Soc, 853 (1963) • 17H. H. Strain, M. R. Thomas, and J. *E. Rabinow itch and Govindjee, Photo- J. Katz, Biochim. Biophys. Acta 7_j>> synthesis, John Wiley and Sons, Inc., 306 (1963). New York, 1969, pp. 102-119- X*W. Oettmeier, T. R. Janson, M. C. 2R. K. Clayton, Molecular Physics in Thurnauer, L. L. Shipman, and J. J. Photosynthesis, Blaisdell Publishing Katz, J. Am. Chem. Soc. 8_[, 339 (1977). Co., New York, 1969. 1»H. Brockmann, Jr., W. Trowitzsch, kk Bulletin of Magnetic Resonance and V. Wray, Org. Magn. Reson. 8, 380 unpublished results. (1976). 40P. H. Hynninen, M. R. Was ielewski, 20J. J. Katz, H. H. Strain, A. L. and J. J. Katz, Acta Chem. Scand. Ser Harkness, M. H. Studier, W. A. Svec, T. B33, 637 (1979). R. Janson, and B. T. Cope, J. Am. Chem. 4°aP. H. Hynninen and G. Sievers, Soc. 2i», 7938 (1972). Z. Naturforsch. Teil B ^6, 1000 21R. C. Dougherty, H. H. Strain, W. (1981) . A. Svec, R. A. Uphaus, and J. J. Katz, 41P. A. Ellsworth and C. B. Storm, J. J. Am. Chem. Soc. 88, 5037 (1966). Org. Chem. k$, 28l (1978). 22R. C. Dougherty, H. H. Strain, W. 42R Willsiatter and A. Stoll, Inves- A. Svec, R. A. Uphaus, and J. J. Katz, tigations on Chlorophyll (trans, by F. J. Am. Chem. Soc. %2, 2826 (1970). M. Schertz and A. R. Merz), Science 23H. H.Strain, B. T. Cope, G. N. Printing Press, Lancaster, Pennsylva- McDonald, W. A. Svec, and J. J. Katz, nia, 1928, p. 131. Phytochemistry K), 1109 (1971). 43J. J. Katz, R. C. Dougherty, F. C. 24H. Scheer, W. A. Svec, B. T. Cope, Pennington, H. H. Strain, and G. L. M. H. Studier, R. G. Scott, and J. J. Closs, J. Am. Chem. Soc. 8_5_, 4049 Katz, J. Am. Chem. Soc. 9_6, 3714 (1963) (197*0 • 44D. Mauzerall and A. Chivis, J. 25A. S. Holt, in The Chlorophylls, L. Theor. Biol. 42, 387 (1973)• 45 P. Vernon and G. R. Seely, Eds., Aca- J. Franck, J. L. Rosenberg, and C. demic Press, New York, 1966, pp. Wei ss in Luminescence of Organic and 111-118. Inorganic Materials, H. P. Kallmann and 2'H. Brockmann, Jr., A. Gloe, N. G. M. Spruch, Eds., John Wiley and Risch, and W. Trowitzch, Liebiqs Ann., Sons, New York, 1962, 4 566 (1976). 'H. Scheer and J. J. Katz, J. Chem. 27N. Risch, T. Kemmer, and H. Brock- Soc. 22, 3273 (1975)• 47 mann, Jr., Liebigs Ann., 585 (1978). H. Scheer and J. J. Katz, J. Am. 28A. S. Holt, J. W. Purdie, and J. W. Chem. Soc. 100, 56I (1978). 48 F. Wasley, Can. J. Chem. 44, 88 A. A. Krasnovskii, Dokl Akad. Nauk SSSR 60, 421 (1948). (1966). 4 2'H.-C. Chow, M. B. Caple, and C. R. 'A. A. Krasnovski i in Progress in Strouse, J. Chromatog. 151. 357 Photosynthetic Research, H. Metzner, (1978). Ed., International Union of Biological 30N. Risch, H. Brockmann, Jr., and A. Sciences, Tubingen, 1969, Vo 1 . 2, pp. Gloe, Liebigs Ann., i*08 (1979). 709-727. 31T. Kemmer, H. Brockmann, Jr., and 50H. Scheer and J. J. Katz, Proc. N. Risch, Z. Naturforsch. Teil B _3_4, Natl. Acad. Sci, USA 21• i626 (1974). 633 (1979) • 51D. N. Lincoln, V. Wray, H. Brock- 32W. Trowitzch, Org. Magn. Reson. 8, mann, Jr., and W. Trowitzsch, J. Chem. 59 (1976) . Soc. Perkin I I, 1920 (1974) . 33J. K. M. Sanders, J. C. Waterton, 52K. M. Smith and J. F. Unsworth, and I. S. Denniss, J.Chem. Soc. Perkin Tetrahedron 3J, 367 (1975) • I, 1150 (1976). 53T. R. Janson and J. J. Katz, Ann. 34J. K. M. Sanders, Chem. Soc. (Lon- N. Y. Acad. Sci. 206, 579 (1973). don) Rev. 6, 467 (1977) . 54J. J. Katz, T. R. Janson, A. G. 3SH. H. Strain and W. M. Manning, J. Kostka, R. A. Uphaus, and G. L. Closs, Bioi. Chem. 146, 275 0942). J. Am. Chem. Soc. 94, 2883 (1972). 3'J. J. Katz, G. D. Norman, W. A. 55L. Shipman, T. R. Janson, G. J. Svec, and H. H. Strain, J. Am. Chem. Ray, and J. J. Katz, Proc. Natl. Acad. Soc. 20, 6841 (1968). Sci USA 22, 2873 (1975). 37F. C. Pennington, H. H. Strain, W. 5'S. G . Boxer, G. L. Closs, and J. J. A. Svec, and J. J. Katz, J. Am. Chem. Katz, J . Am. Chem. Soc. 36, 7058 Soc. 86, 1418 (1964) . (1974). 3"P. H. Hynninen, Acta. Chem. Scand. 57T. R Janson and J. J. Katz in The 22, 1487 (1973). Porphyrins, D. Dolphin, Ed., Academic 3'A Ault, W. A. Svec, and J. J. Katz, Press, New York, 1979. Vol. 4B, pp.

Vol. 5, No. 1/2 45 1-59. Biophys. Acta 408, 349 (1975)- S8C. E. Strouse, V.H. Kollman, and N. 75T. Wydrznski, N. Zumbulyadis, P. G. A. Matwiyoff, Biochem. Biophys. Res. Schmidt, H. S. Gutowsky, and Govind- Commun. 46, 328 (1972). jee, Proc. Natl. Acad. Sci. USA 21. S'N. A. Matwiyoff and B. F. Burnham, 1196 (1976). Ann. N. Y. Acad. Sci. 206, 3&5 (1973). 75aH. H. Robinson, R. R. Sharp, and t0R. A. Goodman, E. Oldfield, and A. C. F. Yocum, Arch. Biochem. Biophys. Allerhand, J. Am. Chem. Soc. 9Ji, 7553 207, 1 (1981). (1973). 7'J. J. Katz, Dev. Appl. Spectr. 6, 41K. M. Smith, M. J. Bushel], J. Rim- 201 (1968) . mer, and J. F. Unsworth, J. Am. Chem. 77J. J. Katz, H. H. Strain, D. L. Soc. 102, 2437 (1980). Leussing, and R. C. Dougherty, J. Am. *2V. Wray, U. Jurgens, and H. Brock- Chem. Soc. 9.0. 784 (1968) . mann, Jr., Tetrahedron _3_5_, 2275 (1979). 7"V Georghe, L. Tuglea, C. Balanescu, '2aS. Lotjonen and P. H. Hynninen, and V. Simplaceanu, Rev. Roum. Phys. Org. Magn. Reson. J_6, 304 (1981) . 21, 1179 (1978). "R. C. Dougherty, G. D. Norman, and 78J. R. Larry and Q. VanWinkle, J. J. J. Katz, J. Am. Chem. Soc. 87, 58OI Phys. Chem. 21. 570 (1969). (1965) • 80J. J. Katz and H. J. Crespi, Pure i4R. C. Dougherty, H. H. Strain, and Appl. Chem. 12, 221 (1972) . J. J. Katz, J. Am. Chem. Soc. 87, 104 81J. J. Katz, G. L. Closs, F. C. Pen- (1965) • nington, M. R. Thomas, and H. H. <5J. J. Katz, R. C. Dougherty, W. A- Strain, J. Am. Chem. Soc. 8_£, 38OI Svec, and H. H. Strain, J. Am. Chem. (1963) • Soc. 86, 4220 (1964). a2F. K. Fong and V. J. Koester, J. "J. R. Norris, R. A. Uphaus, H. L. Am. Chem. Soc. 9_7_. 6888 (1975). Crespi, and J. J. Katz, Proc. Natl. 83F. K. Fong, J. Am. Chem. Soc. 9_2, Acad. Sci. USA 68, 625 0971). 6890 (1975) • •7J. R. Norris, H. Scheer, M. E. Dru- 84L. L. Shipman, T. M. Cotton, J. R. yan, and J. J. Katz, Proc. Natl. Acad. Norris, and J. J. Katz, J. Am. Chem. Sci. USA 21- J+897 (1974) . Soc. 9jS, 8222 (1976) . "H. Scheer, J. J. Katz, and J. R. 85H. Scheer, unpublished results. Norris, J. Am. Chem. Soc. 9_9_, 1372 "A. Rasquain, C. Houssier, and C. (1977). Sironval, Biochim. Biophys. Acta 462, "J. K. M. Sanders and J. C. Water- 622 (1977)• ton, J. Chem Soc. Chem. Commun. 247 87A. D. Trifunac and J. J. Katz, J. (1976). Am. Chem. Soc. 9j>. 5233 (1974). 70J. C. Waterton and J. K. M. San- 88J. R. Norris, R. A. Uphaus, H. L. ders, J. Am. Chem. Soc. 100, 4044 Crespi, and J. J. Katz, Proc. Natl. (1978)> Acad. Sci. USA 68, 625 (1971). 7oaR. G. Brereton and J. K. M. San- 8'J. J. Katz and J. R. Norris, Curr. ders, Org. Magn. Reson. J_9_, 150 Top. Bioenergetics 5_, 41 (1973) - (1982) . '°F. K. Fong, Proc. Natl. Acad. Sci. 71G. L. Closs and E. V. Sitzmann, J. USA 21. 3692 (1974). Am. Chem. Soc. K>3_, 3217 (I98I) . '*S. G. Boxer and G. L. Closs, J. Am. 7xaS. G. Boxer and G. L. Closs, J. Chem. Soc. 9j. 5406 (1976). Am. Chem. Soc. 9J, 3268 (1975). '2L. L. Shipman, T. M. Cotton, J. R. 72T. R. Janson, A.G. Kostka, S. R. Norris, and J. J. Katz, Proc. Natl. Litteken, and J. J. Katz, in Porphyrin Acad. Sci. USA 21. 1791 (1976). Chemistry Advances, F. R. Longo, Ed., '3J. J. Katz, J. R. Norris, and L. L. Ann Arbor Science Publishers, Ann Shipman, Brookhaven Symp. Biol. 2J3, 16 Arbor, Michigan, 1979, PP> 68-70. (1976). 73T. J Wydrzynski, S. B. Marks, P. G. '4M. R. Wasielewski, M. H. Studier, Schmidt, Govindjee, and H. S. Gutowsky, and J. J. Katz, Proc. Natl. Acad. Sci. Biochemistry 12. 2155 (1978). USA 21. I»282 (1976) . 7*T. Wydrzynski, N.Zumbulyadis, P. G. '5M. R. Wasielewski, U. H. Smith, B. Schmidt, and Govindjee, Biochim. T. Cope, and J. J. Katz, J. Am. Chem.

46 Bulletin of Magnetic Resonance Soc. 22, 4172 (1977). 344 (1965). "S. G. Boxer and R. R. Bucks, J. Am. 115M Akhtar, Comp. Biochem. Physiol. Chem. Soc. JOJ, 1883 (1979). 28, 1 (1969). *7J. J. Katz and H. L. Crespi in 114H. A. 0. Hill, J. M. Pratt, and R. Recent Advances in Phytochemistry, M. J. P. Williams, Methods Enzymol. Jj, 5 K. Seikel and V. C. Runeckles, Eds, (197D. Appleton- Century-Crofts, New York, 117G. N. Sando, R. L. Blakely, H. P. 1969, Vol 2, pp. 1-34. C. Hogenkamp, and P. J. Hoffmann, J. "R. C. Dougherty, H. L. Crespi, H. Biol. Chem. 25J), 8774 (1975) • H. Strain, and J. J. Katz, J. Am. Chem. 11$J. R. Pilbrow in B12, EPR of B12- Soc. 88, 2854 (1966). Dependent Enzyme Reactions and Related "J. J. Katz, R. C. Dougherty, H. L. Systems, D. Dolphin, Ed., Wileylnter- Crespi, and H. H. Strain, J. Am. Chem. science, New York, 1982, Vol. 1, pp. Soc. 88, 2856 (I966). 1 431-462. ° ° E •A. Ahrens, Jr., D. C. Williams, X1'T. E. Walker, H. P. C. Hogenkamp, and A. R. Battersby, J. Chem. Soc. T. E. Needham, and N. A. Matwiyoff, Perk ins I . 2540 (1977). Biochemistry JJ, 2650 (1974) 101 R. Bonnett, Chem. Rev. 6j, 573 120D. R. Eaton and W. D. Phillips, (1963) • Advan. Magn. Reson. J, 103 (1965) 102D. D. Hensens, H. A 0. Hill, J. 121G. N. Schrauzer and L. P. Lee, J. Thornton, A. M. Turner, and R, J. P. Am. Chem. Soc. 9J), 6541 (1968) . Will iams, Phi 1. Trans. R Soc. Lond (B) 122J. H. Bayston, F. C. Looney, J. R. 273, 353 (1976). Pilbrow, and M. E. Winfield, Biochemis- 103A. Eschenmoser, E Bertele, H. try 2. 2164 (1970). Boos, J. D. Dunitz, F. Elsinger, I . 123W. Friedrich, Z. Naturforsch, Teil Felner, H. P. Gribi, H. Gschwend, E, F. B 2J., 138 and 595 (1966) . Meyer, M. Pesaro, and R. Scheffold, 12*R. A. Firth, H. A. 0. Hill, B. E. Angew. Chem. 7_6, 393 (1964). Mann, J. M. Pratt, R. G. Thorp, and R. 104A Eschenmoser, R. Scheffold, E. J. P. Williams, J. Chem. Soc. Ser. A, Bertele, M. Pesaro, and H. Gschwend, 2419 (1968). Proc. Roy. Soc. A288, 306 (1965)• 125T. E. Needham, N. A. Matwiyoff, T. 105R. Bonnett and P. G. Redman, Proc. E. Walker, and H. P. C. Hogenkamp, J. Roy. Soc. A288, 342 (1965) • Am. Chem. Soc. 9_£» 5019 (1973). 1O 12< 'H. A. 0. Hill, J. M. Pratt, and R. D. Doddrell and A. Allerhand, J. P. Wi11i ams, J Chem. Soc, 2859 Proc. Natl. Acad. Sci. USA 68, IO83 (1965). (197D • 107 H. A. 0. Hill, B. E. Mann, J. M. 127D. Doddrell and A. Allerhand, J. Pratt, and R. J. P. Wi11iams, J. Chem. Chem. Soc. Chem. Commun., 728 (1971). Soc. (A) , 564 (1968) . 128J. W. Corcoran and D. Shemin, Bio- 108 S. A. Cockle, H. A. 0. Hill, R. J. chim. Biophys. Acta 2^, 66l (1957). P. Williams, B. E. Mann, and J. M. 12'R. Bray and D. Shemin, Biochim. Pratt, Biochim. Biophys. Acta 2_J5_, 415 Biophys. Acta ^0, 647 (1958). (1970) . 130R. C. Bray and D. Shemin, J. Biol. 1O 'P. Y. Law, D. G. Brown, E. L. Chem. 2_3_8, 1501 O9&3) • Lien, B. M. Babior, and J. M. Wood, 131C. E. Brown, J. J. Katz, and D. Biochemistry JO, 3428 (1971). Shemin, Proc. Natl. Acad. Sci. USA 6°,, 110 W. D. Phillips, M. Poe, J.F . Wei- 2585 (1972). her, C. C. McDonald , and W. Levenberg, 132A. I. Scott, C. A. Townsend K. Nature (London) 2_2J, 574 (1970). Okada, M. Kajiwara, P. J. Whitman, and 11X M Poe, W. D. Phillips, C. C. McDo- R. J. Cushley, J. Am. Chem. Soc. 9_4_> nald, and W. Lovenberg, Proc. Natl. 8267 (1972). Acad. Sci. USA |5_, 797 (1970) . 133A. I. Scott, C. A. Townsend, K. 112 J. D. Brodie and M. Poe, Biochem- Okada, M. Kajiwara, and R. J. Cushley, istry JO, 914 (1971). J. Am. Chem. Soc. $h, 8269 (1972). 113 J. D.. Brodie and M. Poe, Biochem- 134A. I. Scott, C. A. Townsend, and istry J_J, 2534 (1972) . R. J. Cushley, J. Am. Chem. Soc. 9_5_. l F. Wagner, Proc. Roy. Soc. A288, 5759 (1973). Vol. 5, No. 1/2 47 X35C. E. Brown, D. Shemin, and J. H. P. C. Hogenkamp, Biochemistry 13, J.Katz, J. Biol. Chem. 2l»8, 8015 26*5 (197*)• (1973). 154H. Johansen and L. I. Ingraham, J. 13'R. Bonnett, J. M. Godfrey, and Theor. Biol. 2J, 191 (1969). V.B. Math, J. Chem. Soc. Ser. C, 3736 155G. N. Schrauzer, L. P. Lee, and J. (197D. W. Silbert, J. Am. Chem. Soc. 22. 2997 (1970) . 137B. V. Cheney and D. M. Grant, J. 15«P. Offenhartz, B. H. Offenhartz, Am. Chem. Soc. 82, 5319 (1967). and M. M. Fung, J. Am. Chem. Soc. 92, 13«D. K. Dalling and D M. Grant, J. 2967 (1970). Am. Chem. Soc. 2*' 5318 (1972). 13 1S7N. A. Matwiyoff and R. S. Drago, 'A. R. Battersby, M. lhara, E. J. Chem. Phys. 18, 2583 (1963). McDonald, J R. Stephenson, and B. T. Golding, J. Chem. Soc. Chem. Commun. 158N. Muller and D. E. Pritchard, J. kOk (1973). Chem. Phys. £[, 11*71 (1959) • 140A. R. Battersby, M. lhara, E. 1S'P Day, Theor. Chim , Acta 2> 328 McDonald, and J R. Stephenson, J. (1957) Chem. Soc. Chem. Commun., i+58 (197*) • 1 6 Op Day, Coord. Chem. Rev. 2, 109 141M. Imfield, C. A. Townsend, and D. (1967) Ar igoni, J. Chem Soc. Chem. Commun., 1 i ip L. Lenhert, Proc R. Soc. Lond. 5*1 (1976). Ser. A 303, *5 (1968). 142A. Battersby R. Hollenstein, E. 1 6 2 p A. Mil ton and T. L. Brown, J. McDonald, and C. Williams, J. Chem. Am. Chem. Soc. 22. 1390 (1977). Soc. Chem. Commun. 5*3 0976) . X'3D. Gust, R. B. Moon, and J. D. 143A. I. Scott, M.Kajiwata, T. Tak- Roberts, Proc. Natl. Acad. Sci . USA 7_2, ahashi, I. A. Armitage, P. Demou, and 1*696 (1975) . D. Petrocine, J. Chem. Soc. Chem. Com- 164J. D. Satterlee Biochem. Biophys. mun. , 5** (1976) Res. Commun. 8, 272 (1979). 144R. Deez, H. P. Kriemler, K. H. It5J. D. Satterlee, Inorg. Chim. Acta Bergmann, and G. Muller, Hoppe-Seyler's *6. 157 (1980). Z Physiol. Chem. 358, 339 (1977). 16'P. K. Mishra, R. K. Gupta, P. C. 145 A. I. Scott A. J. I rwi n, L. M. Goswami, P. N. Venkatosubramanian, and Siegel, and J. N.Shoolery, J. Am . Chem. A. Nath, Biochim. Biophys. Acta 668, Soc. JOO, 7987 (1978). 1*06 (1981). 14'A.R. Battersby, E.McDonald, H. R. 147M. Katada, S. Tyagi, A. Nath, R. Morris, M. Thompson, D. C. Williams, L. Petersen, and R K. Gupta, Biochim. V. Ya Bykhovsky, N. I. Zaitseva, and V. Biophys. Acta 58^, 1*9 (1979)• N. Bukin, Tetrahedron Lett., 2217 1<8H. A. 0. Hill K. G. Moral lee, G. (1977). Costa, G. Pellizer, and A. Lowenstein, 147G. Mueller, K. D. Gneuss, H. P. in Magnetic Resonances in Biological Kriemler, A. I. Scott, and A. J. Irwi n, Research, Magnetic Resonance Studies of J. Am. Chem. Soc. 101, 3655 (1979). Model Molecules of Vi tami n B12, C. 14 "A. Eschenmoser, Chem. Soc. (Lon- Franconi, Ed., Gordon and Breach Sci- don) Rev. 5_, 377 (1976) • ence Publi shers, New York, 1971, PP- 14'A. I. Scott, Accts. Chem. Res. 11 301-309. 1 29 (1978). »»T. L. Brown, Accts. Chem. Res 2. 150A. R. Battersby and E. McDonald, *08 (197*) • Accts. Chem. Res. J_2 . 1* (1979) 170A. Eschenmoser and C. Wi ntner, 15°aA. I. Scott, Pure Appl. Chem. Science 126, 11*10 (1977) • 51. 1215 (1981). 171G. Schlingmann, B. Dresnow, V. l51T. E. Walker, H. P. C. Hogenkamp, Koppenhagen, and L. Ernst, Liebigs Ann. T. E. Needham, and N. A. Matwiyoff, J. Chem., 1186 (I98O). Chem. Soc. Chem. Commun., 85 (197*). X72D. L. Anton, H. P. C. Hogenkamp, 152H. P. C. Hogenkamp, R. E. Tka- T.E. Walker, and N. A. Matwiyoff, J. chuck, M. E. Grant, R. Fuentes, and N. Am. Chem. Soc. lp_2, 2215 (1980) . A. Matwiyoff, Biochemistry jj», 3707 X73D. L. Anton, Diss. Abstr. Int. B. (1975) • hi. 921 (1980). 153R. E. Tkachuck, M. E. Grant, and I740. D. Hensens, H. A. 0. Hill, C.

Bulletin of Magnetic Resonance E. McClelland, and R. J. P. Williams, Charbonneau, S. G. Smith, J. M. Wood, in B12, The Nuclear Magnetic Resonance H. A. 0. Hill, J. M. Pratt, S. Rids- Spectroscopy of Cobalamins and Their dale, and R. J. P. Williams, Biochim. Derivatives, D. Dolphin, Ed., Wiley-ln- Biophys. Acta 304, 85I (1973). terscience, New York, 1982, Vol. 1, pp. I87H. P. C. Hogenkamp, N. A. Kohl- 463-500. miller, R. Hausinger, T. Walker, annd N. 17S L. Ernst, Liebigs Ann. Chem., 376 A. Matwiyoff, Fed. Proc. 3j. 643 (198D. (1979). 17 'V. B. Koppenhagen, B. Elsenhaus, 18>R. T. Taylor, J. A. Happe, M. L. F. Wagner, and J. J. Pfiffner, Fed. Hanna, and R. Wu, J. Environ Sci. Proc. 13_, 1508 (197^) • Health Part A. Environ. Sci Eng, 14, X77Y. Murakami, K. Sakato, Y. Tanaka, 87 (1979). and T. Matsuo, Bull. Chem. Soc. Jpn. l8'Y. Murakami Adv. Chem. Ser. 121, 48, 3622 (1975). 178 179 (1980), J. I. Toohey, Proc. Natl. Acad. X 9 0 LJ C Hull, L. Mauck and B M. Sci. USA 5J». 934 (1965) • Babior, J, Biol Chem. 250, 8023 17 'J. I. Toohey, Fed. Proc. 25_, 1628 (1975) • (1966). 1 9 1 A I Scott and K. Kang, J. Am. 180B. Dresow, G. Schlingmann, L. Chem. Soc. 22. 1997 (1977). Ernst and V. B. Koppenhagen, J. Biol. l»2A. I Scott, J. Kang, D. Dalton, Chem. 25_£, 7637 0980) . and S. K. Chung, J. Am. Chem. Soc. X81B. Dresow, L. Ernst, L. Grotjahn, 100, 3603 (1978) and V. B. Koppenhagen, Angew. Chem. X'3H. Flohr, W. Pannhorst, and J. Int. Ed. Engl. 20, 1048 (I98I) . Retey, Helv. Chim. Acta G 182G. Schlingmann, B. Dresow, L. (1978). Ernst, and V. B. Koppenhagen, Liebigs X'*A. Gaudemer, J. Zylber, N. Zylber, Ann. Chem., 2061 (1981). M. Baran-Marszac, W. E. Hull, M. Foun- 183W. H. Pailes and H. P. C. Hogen- toulakis, A. Konig, K. Wolfle, and J. kamp, Biochemistry 2» 4160 (1968) . Retey, Eur. J. Biochem. 119, 279 184P. Y. Law and J. M. Wood, J. Am. (1981). Chem. Soc. 9J5, 914 (1973) • 1>5A. Cheung, R. Parry, R. H. Abeles, 185 J. Am. Chem. Soc. 102, 384 (1980). P. Y. Law and J. M. Wood, Biochim. X Biophys. Acta ,3_3_i. '•SI (1973). "K. Hikichi and M. Ohuchi, JEOL 18'R. E. DeSimone, M. W. Penley, L. News 18A, 2 (1982).

Vol. 5, No. 1/2 49