T H E JOURNAL OF RIOLOCICAL CHEMISTKY IC) 1991 hy The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 266, No. 22, Issue of August 5, pp. 14440-14445.1991 Printed in U. S. A.

Estimation of Standard Gibbs Energy Changes of Biotransformations*

(Received for publication, February 8, 1991)

Michael L. Mavrovouniotis* From the Systems Research Center and Department of Chemical Engineering, A . V. Williams Building, University of Maryland, College Park, Maryland 20742

Contributions and corrections for the estimation of standard Gibbs energies are given. The group contri- bution method, applicable to both cyclic and acyclic compounds, permits the approximate estimation of the standard Gibbs energy of a biotransformation, given the stoichiometry and structures of the metabolites involved. Estimated standard Gibbs energies of for- mation for a number of acyclic biochemical compounds are provided.

The biological standard Gibbs energy change A@' and the equilibrium constant K', related by the equation

AG"' = -RT 1nK'

are common tools for the thermodynamic analysis of biotrans- formations. Throughout this article, these quantities will refer to concentrations rather than activities (Interunion Commis- sion on Biothermodynamics, 1976), expressed in M (mol/liter or kmol/m"). The standard state is a dilute aqueous solution at pH = 7 and 25 "C. One point requiring particular attention is that the concentration of water is not included (i.e. it is taken equal to 1) in the expression of the equilibrium constant K'. The same is true, at pH = 7, for the concentrations [H"] = 10" and [OH"] = 10". For other values of the pH, the value of K' must be adjusted, based on the stoichiometric coefficients of H" and OH" in the transformation; alterna- tively, one may use the same numerical value for K' but include in its expression the factors ( [H'+]/10-7)" and ([OH"]

where n and m are, respectively, the stoichiometric coefficients of H" and OH" in the transformation.

The quantities AGO' and K' depend on the nature of the biochemical compounds involved and the stoichiometry of the transformation, but they are independent of the kinetics of the enzyme or enzymes catalyzing the transformation.With u, the stoichiometric coefficient of compound Si, a transforma- tion with known stoichiometry can be written as shown below,

u s , = 0

with the usual sign convention for u,, specifically u, > 0 if S, is a net product and u, < 0 if Si is a net reactant or substrate. If AG!" is the Gibbs energy of formation of Si, the standard

* This research was supported in part by National Science Foun- dation Grant CTS 9010549 and the University of Maryland Systems Research Center which is also funded by National Science Foundation Grant CDR 8803012. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore he hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ T o whom correspondence should be addressed. Tel.: 301-405- 6620; Fax: 301-405-6707; Internet electronic mail: mlmavro@ phoenix.src.umd.edu.

Gibbs energy change of the transformation can be expressed as shown below.

AG"' = 1 u;AGY'

Thus, the standard Gibbs energies of formation of a number of biochemical compounds allow the determination of the standard Gibbs energy change (and the equilibrium constant) of any transformation that involves only these biochemical compounds. It was previously mentioned that the concentra- tions of water (regardless of pH) and the concentrations of H" and OH" (for pH = 7) are not included in the expression for K'; it should be emphasized here that the Gibbs energies of formation of these species must nonetheless be included in the derivation of the Gibbs energy change of the transforma- tion. In the rest of this article, the term "Gibbs energy" will be used for both the standard Gibbs energy of formation of a compound and the standard Gibbs energy change of a trans- formation.

For most compounds and biotransformations, Gibbs ener- gies are not available. Various thermodynamic properties can often be estimated from the structures of the compounds involved, using group contributions (Benson, 1968; Reid et al., 1987). This article presents group contributions and new results on an group contribution estimation approach, whose basic characteristics have been discussed in a previous publi- cation (Mavrovouniotis, 1990a). The method is based on multiple linear regression, using data from several sources (Barman, 1969a; 1969b; 1974; Thauer et al., 1977; Hinz, 1986; Lehninger, 1975; 1986; Sober, 1970; Edsall and Gutfreund, 1983), screened and weighted according to their accuracy. Theoretically, the method could be used for any organic compound in aqueous solution, but it may be less accurate for non-biochemical compounds, because the data from which the contributions were estimated were very heavily biased in favor of biochemical compounds and reactions.

THEORY'

Relating Biotransformations to Groups-In group contribu- tion methods, one views a compound as composed of func- tional groups and sums amounts contributed by each group to the Gibbs energy of formation of the compound. Correc- tions for special structural features of the compound can also be added to the Gibbs energy. The contributions of groups or features that occur more than once in the structure of a compound must be multiplied by the number of occurrences. Thus, the Gibbs energy AG? of a compound Si is given by an expression of the form shown in the following equation,

Portions of this paper (including Tables I-V and Figs. 1-3) are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are included in the microfilm edition of the Journal that is available from Waverly Press.

14440

Gibbs Energy Changes of Biotransformations 14441

AGP' = 2 milgj

where mi, is the number of occurrences of group j in Si, and g, is the contribution of group j . For the Gibbs energy of for- mation (as well as other thermodynamic quantities, such as the standard enthalpy of formation), these methods entail only linear combination, without any other functional trans- formation. For the Gibbs energy A@' of a biotransformation,

2 uisi = 0,

I

one can combine the expressions AGO' = u~AGP'

and

A@' = C m,g, J

t o obtain the following.

AG"' = vi m,Jg, I J

* AG"' = (1 u,miJ)gJ J I

* AGO' = n,gJ, J

where

n, = C u,m,,

The coefficient nj represents the net participation of group (or correction) j in the biotransformation; it is obtained by subtracting the number of occurrences of each group in reac- tants from its occurrences in products, always taking into account the stoichiometric coefficients. Thus, the estimation of the Gibbs energy of a transformation from contributions of groups can be carried out directly, without the intermediate estimation of the Gibbs energies of the participating com- pounds. A group that occurs the same number of times in reactants and products leads to a corresponding nj = 0; thus, it does not affect the Gibbs energy of the transformation and its contribution is not necessary. Often, a transformation involves complicated reactants and products but affects only a small portion of their structures; only this affected portion must be decomposed into groups for the estimation of the Gibbs energy. The way in which group contribution methods operate will be clarified further in the context of an example, after the detailed group contributions are presented.

Group Contribution.!-The groups and other features used in the method, along with their contributions to the Gibbs energy, are listed in Table I. The contributions were estimated through multiple linear regression, based on literature data (Barman, 1969a; 1969b; 1974; Thauer et al., 1977; Hinz, 1986; Lehninger, 1!375; 1986; Sober, 1970; Edsall and Gutfreund, 1983). Throughout the application of the group contribution method, all compounds and groups must be represented in their common state in aqueous solution, at the standard state of pH = 7 and temperature 25 "C. For example, amine groups should normally have a proton attached, as in R-NH,i+ rather than R-NH2, carboxylic acids should be in their anion form R-CO-0" rather than R-CO-OH, amino acids in their zwit- ter-ion forms, etc. I t should be pointed out in particular that the proposed method cannot be used to estimate acid-base equilibria (e.g. the K, dissociation constant for a carboxylic acid or a protonated amine), because the method applies only to the most prevalent of the acid-base forms of a compound.

In Table I, one encounters groups that at first appear to permit such acid-base calculations (for example, both -NH2 and -NHi+ are listed); the contribution of each form of a group, however, presupposes a biochemical compound struc- ture that makes that form prevalent at pH = 7.

In order to apply the developed group contribution tech- nique, one must determine which groups or features occur in the structure of the compound (and how many times, in the case of multiple occurrences). In this determination, one must always use the most specialized possible group. The arrange- ment of the groups in Table I is designed to facilitate this.

The first contribution in Table I refers to the origin, which is used as the initial quantity for the Gibbs energy of any biochemical compound. The be treated like any other group in the estimation of the Gibbs energy of the transformation; one origin term, multiplied by the appropriate stoichiometric coefficient, must be used for each participating compound. Thus, for a biotransformation of either the form A + B + C or the form A + 2B, the net effect on A@' is the addition of one origin term; for a transformation of the form A + D + B + C the net effect is zero; while for a transformation of the form A + 2D + C the net effect is subtraction of two origin terms.

The next set of contributions in Table I refers to the contributions of features that are not represented by distinct functional groups. When the Gibbs energy of a transformation is sought, the inclusion of these features and contributions must be treated according to the stoichiometry of the trans- formation, in a way similar to the treatment of the origin.

The presence of one of these features in a compound and the inclusion of the appropriate contribution do not eliminate any portion of the molecule from further consideration. For example, if a molecule consists of only carbon and hydrogen, the hydrocarbon correction of 4.0 kcal/mol must be used in addition to the contributions of the groups that make up the structure of the molecule. Similarly, the amide correction of -10.5 kcal/mol must be used for each occurrence of a nitrogen which attached to a carbonyl group, in addition to the contri- butions of the actual groups (from the rest of Table I).

The remainder of Table I refers to specific groups and has been organized so that groups are easy to locate, and the more specialized groups are listed first. In contrast to feature-based corrections, occurrence of a group in a compound eliminates the use of the same atoms of the compound in a subsequent group; as noted earlier, the most specialized possible group should be used, and the order of the groups in the table reflects this priority.

In practice, one focuses first on any occurrences of sulfur and/or phosphorus in a compound. Scanning Table I under the appropriate subheading, starting at the top, one locates the groups that must be used. Each group is marked, essen- tially eliminated from the structure of the compound. One then proceeds similarly to nitrogen atoms, always picking the most specialized group, i.e. the first group that matches the structure (based on the order in which they are provided under the nitrogen subheading). Occurrences of oxygen that have not already been included in other groups (nitrogen or phosphorus) are considered next. Finally, the remaining por- tions of the structure of the compound are broken down, first into unsaturated carbon groups and then saturated carbon groups. In the end, the whole structure must be used up by groups, and if any portion of the structure remains then the method cannot be applied, i.e. the structure is outside the scope of the method. Note, in particular, that there is no group for hydrogen (i.e. -H); if hydrogens remain in the structure not accounted for in groups, it is likely that an

14442 Gibbs Energy Changes of Biotransformations

incorrect group was used in the decomposition. For example, if one incorrectly uses a >C< group instead of a >CH- group, a hydrogen will remain unused.

In Table I, the term “benzene ring” is used for a six-carbon aromatic ring (without heteroatoms). The term “heteroaro- matic ring” refers to a ring which contains a heteroatom and, following Huckel’s rule, has six delocalized *-electrons.

As mentioned earlier, for biotransformations one is con- cerned only with that portion of the compounds which ac- tually undergoes a modification. Not only groups but also feature-based corrections affected by the modification must be considered in the estimation of the transformation’s AG”.

Group contribution methods are generally inaccurate for very small compounds, including inorganic ions, which in fact cannot be broken down into groups, and some organic com- pounds with two or fewer carbons (e.g. methane, formalde- hyde, oxalate), which may or may not be formally decompos- able into groups. The Gibbs energies of these small com- pounds can be obtained either from a reliable source of experimental data (Thauer et al., 1977) or from estimates that were developed in conjunction with group contributions (Mav- rovouniotis, 1990a). The same sources can be used for certain complex biochemical compounds with important metabolic roles (such as the pair NAD’/NADH).

Table I1 provides the Gibbs energies of formation of com- mon acyclic compounds. The Gibbs energies of common cyclic compounds, from an earlier version of the method, have been previously reported (Mavrovouniotis, 1990a). Improvements in the method have led to changes in the values for cyclic compounds; if one computes a Gibbs energy from the contri- butions provided here and compares it with that previously reported, a small difference is likely to be observed. However, the changes are not sufficient to warrant the presentation of the new results for cyclic compounds in this article. Table I1 has been accordingly confined to the estimation results for acyclic compounds, which have never been reported. The energies of Table I1 can serve as useful reference points, when the Gibbs energy of a compound similar to one given in the table is desired. One needs to identify only the differences between the compound at hand and the similar compound listed in the table and add and subtract appropriate contri- butions (and corrections) that correspond to these differences. The Gibbs energies of Table I1 were estimated from group contributions that had not been rounded off; because of this, if one estimates the Gibbs energy for a compound from the contributions of Table I and compares it with a value given in Table 11, a small deviation might be observed in the first decimal digit.

As an example of the application of the method, Fig. 1 depicts the decomposition of the structure of cystathionine. The computation of its standard Gibbs energy of formation is shown in Table 111. As an illustration of the application of the group contribution method to biotransformations, con- sider the example of Fig. 2.

Cystathionine + acetate + homocysteine + pyruvate + acetamide

This transformation cannot be achieved in a single enzy- matic step, but it represents a legitimate overall transforma- tion for which a pathway of enzymatic and/or nonenzymatic steps might exist. What is essential here is that the transfor- mation is stoichiometrically balanced. Expressing the trans- formation in terms of molecular formulae,

C7H1404N2S + C,H:qO:- + C4Hg0,NS + C&OA- + C2H50N

one can easily verify that the balances for the electrical charge and all the elements (C, H, 0, N, S) are satisfied. It should

be emphasized that this is a necessary condition for the estimation of the Gibbs energy of the transformation. One could, naturally, estimate separately the Gibbs energies of formation of the five compounds involved (some of which are also given in Table 11), and then derive A@‘ for the transfor- mation as shown in the following equation.

AGO’ = AG”’(hornocysteine) + AG”(pyruvate) + AGo’(acetamide)

- AG“(cystathi0nine) + AG”’(acetate)

However, one can take advantage of the fact that some portions of the chemical structures do not undergo transfor- mation, leading to a simpler computation. Fig. 3a marks the unmodified portions, and Fig. 3b shows how the remainders of the structures are decomposed into groups. Based on this decomposition, the computation of the transformation’s A@’ is detailed in Table IV. The participation of each group is determined as its total number of occurrences in products and reactants, taking into account the stoichiometric coefficients (which are negative for reactants). Note that, regardless of the cancellation of structures (including all groups in the structure of acetate), the origin contribution must be included with a coefficient of 1, which is equal to the sum of all the stoichiometric coefficients (with the coefficients of reactants being negative). Note also the inclusion of the amide correc- tion.

The examples given above illustrate the mechanics of ap- plying the group contribution method. Some additional ex- amples are presented in Table V, with a comparison between estimated Gibbs energies and those reported in the literature. Deviations of 1-2 kcal/mol, such as the ones in Table V, are typical, but occasional deviations exceeding 5 kcal/mol are possible (Mavrovouniotis, 1990a).

Concluding Remarks-The contributions of Table I allow the estimation of the standard Gibbs energy of formation of a biochemical compound and the standard Gibbs energy and the equilibrium constant of a biotransformation. The typical error of the group contribution approach is less than 2 kcal/ mol but errors higher than 5 kcal/mol can occur (Mavrovou- niotis, 1990a). Thus, the method provides only an approxi- mate value for the Gibbs energy, correct within a few kcal/ mol. Only the rough order of magnitude of the equilibrium constant can be obtained. The method is not accurate enough for those cases in which a precise value of the equilibrium constant is needed. I t is, however, quite useful in assessing the feasibility and reversibility of bioreactions and pathways.

While the method presented here is, in theory, also appli- cable to non-biochemical organic compounds in aqueous so- lution, there is a considerable bias towards biochemical com- pounds in the database from which the contributions were derived. Hence, extreme caution should be used when non- biochemical compounds are considered.

To achieve better accuracy and reduce the number of pa- rameters, alternatives to the group contribution approach are currently being considered, including an approach that bases the estimation on alternative formal arrangements of valence electrons (Mavrovouniotis, 1990b).

REFERENCES

Barman, T. E. (1969a) Enzyme Handbook, Vol. 1, Springer-Verlag,

Barman, T. E. (1969b) Enzyme Handbook, Vol. 2, Springer-Verlag,

Barman, T. E. (1974) Enzyme Handbook, Suppl. 1, Springer-Verlag,

Benson, S. W. (1968) Thermochemical Kinetics, John Wiley & Sons,

New York

New York

New York

New York

Gibbs Energy Changes of Biotransformations 14443 Edsall, J . T., and Gutfreund, H. (1983) Biothermodynamics, John

Hinz, H.-J. (1986) Thermodynamic Data for Biochemistry and Bio-

Interunion Commission on Biothermodynamics (1976) J. Biol. Chem.

Lehninger, A. E. (1975) Biochemistry, 2nd Ed., Worth, New York Lehninger, A. E. (1982) Principles of Biochemistry, Worth, New York

Wiley & Sons, New York

technology, Springer-Verlag, New York

25 1,6879-6885

Supplemental Material to Estimation of Standard Gibbs Energy Changes of Biotransformations

Michael L. Mavrovouniotis

The contrtbut~ons 01 groups which can be used to esl8mate the standard Glbbs Energy change of a blotransformatlon or the standard Gibbr Energy 01 formallon of a blochemlcal compound 8 " aqueous s01ut~on are glven in Table I Compounds and groups must always be represented In thelr Common state m aqueous so1ut1on. at pH=7 and temperature 25% (e g., R-CO-0'- rather than R-CO-OH). In decomposing the structure 01 a Compound loto groups. one musl always use the most specialized possble group; the groups 8n Table I are Ordered 50 that more speclallzed groups occumng earller

Table I lncludes the origm (whlch 1s a contrlbumn added to the Glbbs energy of any blochemlcal compound). and features that are not represented by dstfnct IUnCtlOnal groups The inclusion 01 a pmon 01 a molecular Structure Inlo a group precludes 11s I ~ C I U S I O ~ in other groups. the SpecIaI features however represent corrections that must st811 be Included.

In pract~ce. one 1ocuses llrst on heteroatoms (sulfur. phosphorous. nltragen. and oxygen. rn that order) and ldenthes the necessary groups under the appropriate subheading 01 Table I. each group that is

ldenlllied can be elmmated Irom further cons~derat~on (except for feature-based corrections) Then. one conslders slmllarly unsaturated carbon groups and Ilnally saturated carbon omups It any pon80n 01 the chemlcal structure remams unused then the method was 88ther applmd tncorrectly or is Smply not applicabie to the s t r ~ d u r e in quesllon

Glbbs energies 01 formallon have been precomputed lor many acycllc compounds and are reponed tn Table 11. Glbbs energies tor many cyclic compounds have been reponed (Mavravoun~otls 1990a) within an earlher version 01 the method

Table 111 shows the computamn 01 11s standard Glbbs Energy 01 formation. Flgure 1 shows. as an example. the decampos,hon 01 the structure 01 cystathlontne Into groups and

by taklng advantage 01 the tact that some pon~ans of the chemlcal structures remaln unchanged Table Figures 2 and 3 show how one can ett~c~ently emmate the Gibbs Energy change lor a blotransformallon

IV provldes the corresponding COmputat8on 01 the translormat~on's AG"' The table shows the net number 01 occurrences 01 each group ln products and reactants. taklng mto account the Slolchlomelr~c coeniclents (whlch are negative tor reactants) Note the 1ncIus80n of the orlgln contrlbutlon and the amtde correction

Additional examples ~n Table V allow a comparison between estfmated Glbbs Energles and those reponed ln the ltterature.

Mavrovouniotis, M. L. (1990a) Biotechnol. Bioeng. 36, 1070-1082 Mavrovouniotis, M. L. (1990b) Ind. Eng. Chem. Res. 29, 1943-1953 Reid, R. C., Prausnitz, J. M., and Poling, B. E. (1987) The Properties

Sober, H. A. (ed) (1970) Handbook of Biochemistry, CRC Press,

Thauer, R. K., Jungermann, K., and Decker, K. (1977) Bacteriol. Reu.

of Gases and Liquids, 4th Ed., McGraw-Hill, New York

Cleveland, OH

41,100-180

Tabla I. Contnbutms 01 groups and corrections to the Standard Gibbs energy 01 formatton. A ring is consdered heteroaromatlc 11 11 has, by resonance. 6 delocalized n-electrons (according to HUckeI's 4 n t 2 rule). The term 'banzene nng' refers to any aromatic rlng that mvolves six carbon atoms and no heteroatoms. when the term 'benzene nng" is applied 10 fused nngr. it should be noted that a benzene rlng can be fused to ether a "on-bmzene ring or another benzene rlng. The subgroup CO denotes a carbonyl group. The contributions 01 >C< (participatlng in two non-benzene rings) and X = (pan~apat~ng ~n two Iused nngs: one benzene rlng and one non.benzene nng) are followed by a lolde (-) to mdlcate that these contrlbutlons should be used wlth extra caution, because they have been estimated mdirealy from the contributions 01 other groups.

GROUP OR CORRECTION CONTRIBUTION (kcalimol)

origin (a contribution which must be added to every compound) :or rect ions for var ious features correction for hydrocarbon molecule (i.e., molecule containing only C and H) correction tor each three-atom ring 28.9

4.0

correclm for each amide (i.e., each nitrogen which is anached to a carbonyl group) -1 0.4 correction for each heteroaromatic ring (following Huckel's rule: -5.9

i u l f u r g roups except lor this correction, heteroaromatic rings are treated like regular non-benzene rings)

-so31- -105.8

-s-s- S H 13.4

5.8 s- I hosphorous g roups

9.5

4 -P0~1- - (participating in a ring) 4 0 - 0 P 0 3 2 - -72.5

14.8

4 P O + (primary) -29.5 4 P O + (secondary) -30.0 4 P 0 3 2 - (tertiary) -25.7 -PO+ 4-P02"-0-

9.5

4-PO2"- -29.8

-5.2 l i t rogen groups -N< (participating in two fused rings) >N=" (the double bond and one single bond particlpating in a ring)

18.9

=N- (participating in a ring) 0.4

>NH (participating in a ring) 10.4

-N< (participating in a ring) 9.5

-NHz 7.6

10.3 -NH31+ EN

4.3 14.9

=NH21+ =NH

0.4 13.6

>NHzl+ >NH

6.9

>N- 7.6 6.3

) xygen g roups >NH~+- 8.6

4 -CC- (participating in a nng) >CO (participating in a ring) -27.4

-54.6

4- (participating in a ring) -24.3 X H = O -17.8 d H (attached to benzene ring) -31.8 4 H (primary) -29.3 4 H (secondary) -32.0 4 H (tertiary) - C o o "

-30.5 -72.0

-C@O- -73.6 > c o 4- -22.5

-27.2

Insa tura ted carbon ihydrogen g roups >C= (Participating in two fused benzene rings) >C= (participating in two fused non-benzene rings) 16.8

2.5

4 H = (participating in a benzene ring) X- (participating in two fused rings: one benzene ring and one non-benzene ring) 6.0

8.4 X = (the formal double bond and a formal single bond padicipating in a benzene nng) 1.5

4 H = (participating in a non-benzene ring) >C= (two single bonds participating in a non-benzene ring) 22.8

>C= (the double band and one single bond participating in a non-benzene ring) 8.2 9.6

=CH =C-

35.7

-CH= 24.0

-CH2 11.1

=G 18.4

a t u r a t e d c a r b o n l h y d r o g e n g r o u p s 5.0

>CH- (participating in two (used non-benzene rings) >C< (participating in two non-benzene rings)

4 . 9

>CHz (participating in a non-benzene ring) -12.0-

>CH- (participating in a non-benzene ring) 6.1

>Cc (participating in a non-benzene ring) -12.8 -2.2

4 H 3 K H z

7.9 1.7

>CH- >C< -12.8

4 . 8

-23.6

14444 Gibbs Energy Changes of Biotransformations Tabk ll. Estimated standard Gibbs energles 01 forrnatlon of selected acyclic compounds The compom@s lor which mSUIl6 are gwan below are only those acycllc compounds that Occurred in the database used in t h e course 01 the development 01 this method. The compounds are grouped under subhe.dings, based On key heteroatoms. Wlthin each subheading, they are Icsted ~n descending order of number of nitrogen atoms, then oxygen% then carbon. and finally hydrogens.

Compound Molecular Formula A G O '

(kcawmol)

Sulfur compounds cymn.

methtonins

homocystme

cysteine

d#methyl-Sultide

mhyl-mercaptan

Phosphate compounds EBdoh~plUlOS~-I,7-Lphosphale

3.5-a~pho~phome~alDnats

3-phosphogtyc~royl-ph~sphate

2.3.a~phorphog1ycerste

Isopsntsnyl-pyrophosphate

dlm(lthylallyl-pyl0phosphate

srgmne-phosphate

creallne-phosphate

8.aspanyl-phosphatw

3-phospho-1-serine

carbamoyl-phosphate

~ e d o h e p t ~ l o ~ e - 7 - p h o ~ p h a t e

mann~tol.l.phosphats

6.pho~pho-2-dehydr0-3-d~~~y-glUCOnale

~yI~lo0e-5-ph0~phale

ribulose-5.phorphats

5-phospho-mevalonate

eryfhross-4-pho~phate

glyCerOletr~IOS~t-phosphate

3-phosphoglycerate

2.phosphogly~erate

3-pho~pho-hydroxy-pyluvate

glycerol-3-phosphate

dihydrolyaCBtOne-phoSphatE

glyceraldehyd~-3-phosphate

phoPphoeno1pyrUvate

acetyl-phosphate

Nitrogen compounds arpinlnowcc8nate

arplntna

CltrYlIhn.

creatine

guanidino-acetate

n - s u c c i ~ y i - a - ~ - d ~ a ~ ~ ~ ~ p ~ m e l a t e

sacchampine

n-ca!brmoyl-aspanat@

msao-a-r-di.m~nopmelate

a-e-diamlnoplmelate

n.lorm~m~no-glutamate

camamOyI.oxmate

glutsmine

5-ursdopropmate

l y m e

2.4.Lammopentanoate

Omthine

tammino.plyc4ne

urea

t.t-alh.n.-Lemine

n.a.ruccin~-a-~mmc-c-kalo.plmelate

n-acetyl-gMamate

lofmyl-p(utamMe

2-aminoadlpale

glutamate

hydmxyrnethylssnnm

aspanate

o-am,naadlpale-&semlaldehyde

glutam.te-semialdehyds

2-am~~o-4-o~o-pentoate

allothreontne

a-melhvl-ser#ne

-1 59.4

-75.2

-79.3

-81 .o 1.9

4 . 6

-253.4 -1 50.6

-160.6

-158 2

-23 5

-24.5

430.9 -57.1

-166.9

-1 23.9

-96.1

-253 2 -226.0 -222 4

-179 6

-179 6

-1 55.3 -142.7

-142.9

-160 1

-158 0

-150.6 -1142

-1 06 1

-105.9

-102.2

-88.1

-217.5

67 .7 -121.0

-63.9

-70.6 -266.4

-227.8

-200.7

-163.5 -163.5

-149 9

-152.8 -120.0

-122.2

-85.0

-87.0 -86.7 -61.6 -50.9 -1 1.6

-293.0

-191.0 -179.2 -163.0

-164.7 -1 59.3 -1 66.4

-1 08.8 -1 10.5 -1 13.6 -1 24.9

-123.8

Gibbs Energy Changes of Biotransformations 14445

lanale

mal0nale.semialdehyde

ketopyruvale

glycollate

glyoxalale

palm!lale

caproate

valerate

2.3.bulaned8ol

elhyl.acelale

acetoin

bulyrale

crotonate

1.2-propanedtol

2-melhoxyelhanol

melhyl.acelale

lanaldehyde

propnonale

acrylate

elhylene.glyCO1

2.hydroxy-acelaldehyde

acetate

l.oclano1

d#prOpyl-elher

herylaldehyde

I-penIan01

dle1hyI.elher

I -b~ lano l

methyl-ethyl-ketone

butyraldehyde

1.propanol

msopropanol

dlmelhyl.kelone

ethanol

dlmelhyl-ether

acelaldehyde

-124.4

-111.7

-1 14.8

-1 23.1

-1 13.4

-63.8

-80.8

-82.5

-81.2

-79.6

-71.6

-84.2

-65.5

-79.9

-63.9

-81.3

-70.2

-85.9

-66.1

-78.7

-68.9

-87.6

-32.9 -23.3

-26.6

-38.1

-26.7

-39.8

-33.2

-30.0

41 .5 -44.4

-34.9

-43.2

-30.1

-33.4

Flgure 1. The SINnUre 01 cyslalh~onine (a) and 1 s deCOmposr1ton mlo groups (b) lor the eslmalron 01 01s slandard Gibbs energy of formal8on (Table 111).

Table 111. Calculation 01 the Gibbs energy 01 formalron of CySlalhionlne (Figure 1) Irom conlnbul8ons 01 groups.

Group or Correcllon Number of Occurrences Conlrlbutlon Total Conlrlbutlon (kcallrnol) (kcallmol)

Origin

-72 0 -144 0

3 1.7 5 1

>CH- 2 -4.8 -9.6

Total Glbbs Energy: -153.9

&CO\ cystathionine

CWHFS-CH~"CH~-CH, ,cO-o"

NH?' NH? acetate C H ~

H-S-CH2"CH2-CH, homocysteine NH?

d-CO, NH;,

..... "..,*.* .... , \,...:, .A ............ acetamide {CH +-

homocysteine

Flgure 3. ldenlificalion of subslNclures which are PfeSeNed In a Ifansformalion (a) and decompos#l#on 01 Ihe remamders of lhe SIN~UfBS lnlo groups (b). lor Ihe eslimalion 01 the Glbbs energy (Table IV)

Table IV. Calculal~on of Ihe Gibbs energy 01 Ihe IfanSformaI~On of Flgure 3. lrom conlr8bulwms 01

groups.

IGrouo or Correclion Number 01 Occurrences Conlrlbutlon Total C o n l r l b ~ l l o n ~ (kcal~mol) (kcalimol)

origm 3-2=1 -23 6 -23 6

amsde correcl~on

-SH

-NHz

.co -CH3

4-

-NH31*

-COO"

>CHz

.CH-

I

1

1

2

1

-1

- 1

-1

-1

-I

-10 4

13.4

10.3

-27 2

7 9

9 5

4.3

-72 0

1 7

-4 8

-10 4

134

I 0 3

-54 4

7 9

-9 5

-4 3

72 0

- 1 7

4 8

Total Gtbbs Energy: 4 5

Table V. Compar~son belwean reponed slandard Gmbbs Energmes and lnose eslmaled lrom Conlnb~t~ons of groups. The devlal!On (last column) IS delrned as the absolule value 01 the ddlerence between Ihe Es18rnaled AGO' and the Reponed AG*'

~

Compound or Transformallon Reported dG" Eslimeted ,AGa' Devlallon (kcaumol) (kcawmol) (kcallrnol)

proptonale -86 3' -860 0 3

glycerol

urea

phenylalanme

lanale

- I 14.0' -1156 1 6

-48 7' -51 2 3

-49 5' - 5 0 9 1 4

-1 23 8' -124 59 0 6

or L.allofhreonme aldolase (EC 4.1.2 6 ) 1-allofhreonme acefaldehyde lyase

1.alloIhfeon~ne "i 1.glyclne f acelaldeyde

ATP L-argmme phoSpholraansleraSe or L.argmme kmse (EC 2 7.3 3)

' T h a w era,, I 977

ThlS value differs from the value 01 -124 4 kcallmol cornpuled m Table 11. The discrepancy 1s due IO

Ihe resuhs shown here were cornpuled lrom the rounded conlr~bul8ons of Table I roundoff errors The resulls of Table II used Ihe conlr#bul#ons belore roundolf 10 Ihe l8rsl decmal. while

1 flOm an equll8bnum COnSlanl gwen by Barman (1969b)

5 average 01 IWO values. 3 9 kcallmol and 2 5 kcaumol. derwed lrom equ~hbrurn conslan~s given by Barman (1969a)

Flgure 2. A Sloichiomelticallybalanced lranslormalion. for which Ihe standard Gibbs energy is IO be eslmaled. The estimation procedure IS Shown in Flgure 3 and Table IV.