Residues that influence coenzyme preference in the aldehyde dehydrogenases.

CBI 7249

No. of Pages 15, Model 5G

16 January 2015 Chemico-Biological Interactions xxx (2015) xxx–xxx 1

Contents lists available at ScienceDirect

Chemico-Biological Interactions journal homepage: www.elsevier.com/locate/chembioint 5 6

4

Q1

Residues that influence coenzyme preference in the aldehyde dehydrogenases

7

Q2

Lilian González-Segura a, Héctor Riveros-Rosas b, Adriana Julián-Sánchez b, Rosario A. Muñoz-Clares a,⇑

3

8 9 10 11 1 2 3 5 14 15 16 17 18 19 20 21 22 23 24

a b

Departamento de Bioquímica, Facultad de Química, Universidad Nacional Autónoma de México, México D.F. 04510, Mexico Departamento de Bioquímica, Facultad de Medicina, Universidad Nacional Autónoma de México, México D.F. 04510, Mexico

a r t i c l e

i n f o

Article history: Available online xxxx Keywords: NAD(P)+ specificity Dual-coenzyme specificity Conformational changes Kinetic parameters Structural and multiple sequence alignments Logos analysis

a b s t r a c t To find out the residues that influence the coenzyme preference of aldehyde dehydrogenases (ALDHs), we reviewed, analyzed and correlated data from their known crystal structures and amino-acid sequences with their published kinetic parameters for NAD(P)+. We found that the conformation of the Rossmann-fold loops participating in binding the adenosine ribose is very conserved among ALDHs, so that coenzyme specificity is mainly determined by the nature of the residue at position 195 (human ALDH2 numbering). Enzymes with glutamate or proline at 195 prefer NAD+ because the side-chains of these residues electrostatically and/or sterically repel the 20 -phosphate group of NADP+. But contrary to the conformational rigidity of proline, the conformational flexibility of glutamate may allow NADP+-binding in some enzymes by moving the carboxyl group away from the 20 -phosphate group, which is possible if a small neutral residue is located at position 224, and favored if the residue at position 53 interacts with the Glu195 in a NADP+-compatible conformation. Of the residues found at position 195, only glutamate interacts with the NAD+-adenosine ribose; glutamine and histidine cannot since their side-chain points opposite to the ribose, probably because the absence of the electrostatic attraction by the conserved nearby Lys192, or its electrostatic repulsion, respectively. The shorter side-chains of other residues— aspartate, serine, threonine, alanine, valine, leucine, or isoleucine—are distant from the ribose but leave room for binding the 20 -phosphate group. Generally, enzymes having a residue different from Glu bind NAD+ with less affinity but they can also bind NADP+ even sometimes with higher affinity than NAD+, as do enzymes containing Thr/Ser/Gln195. Coenzyme preference is a variable feature within many ALDH families, consistent with being mainly dependent on a single residue that apparently has no other structural or functional roles, and therefore can easily be changed through evolution and selected in response to physiological needs. Ó 2015 Published by Elsevier Ireland Ltd.

26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50

51 52 53 54 55 56 57 58 59 60

1. Introduction The aldehyde dehydrogenase (ALDH) superfamily includes NAD+- and NADP+-specific as well as dual-coenzyme specificity enzymes, and the mode of binding of the coenzymes has been determined at atomic detail in several of them. In ALDHs the coenzyme binds in a five-stranded open a/b domain—the Rossmann fold [1]—which in ALDHs has differences with the ‘‘classical’’ domain of other pyridine nucleotide-dependent dehydrogenases [2,3]. The study of the structural reasons behind the coenzyme

Abbreviations: ALDH, aldehyde dehydrogenase; hALDH2, aldehyde dehydrogenase 2 from human; PDB, Protein data bank; PaBADH, betaine aldehyde dehydrogenase from Pseudomonas aeruginosa. ⇑ Corresponding author. Tel.: + 52 55 56225276; fax: +52 55 56225329. E-mail address: [email protected] (R.A. Muñoz-Clares).

preference of ALDHs was addressed long ago by site-directed mutagenesis and X-ray crystallography [4–7]. These early studies showed the importance of having a Glu, at a position in the Rossmann fold similar to the acidic residue (Asp or Glu) in other pyridine nucleotide-linked dehydrogenases, for discriminating against NADP(H) by sterically and electrostatically repelling the 20 -phosphate group of their adenosine ribose [5]. In some other NAD+-dependent ALDHs it was found that a Pro in the place of this Glu prevents the binding of the 20 -phosphate group [8], which was allowed in enzymes having a small and neutral residue, such as Thr or Ser [6,7]. But ALDH enzymes that have Glu at this critical position and are able to bind NADP(H) were also found [5,9]. Therefore, in addition to Glu, other residues are involved in determining the coenzyme preference in these enzymes. In order to find out these residues and to learn more about how the enzymes of this important superfamily bind the coenzymes, here we review,

http://dx.doi.org/10.1016/j.cbi.2014.12.039 0009-2797/Ó 2015 Published by Elsevier Ireland Ltd.

Q1 Please cite this article in press as: L. González-Segura et al., Residues that influence coenzyme preference in the aldehyde dehydrogenases, Chemico-Biological Interactions (2015), http://dx.doi.org/10.1016/j.cbi.2014.12.039

61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76

CBI 7249


16 January 2015 2

L. González-Segura et al. / Chemico-Biological Interactions xxx (2015) xxx–xxx

79

analyze, summarize and correlate the available crystallographic, amino acid sequence and kinetic data of a significant number of ALDH enzymes.

80

2. Methods

77 78

81 82 83 84

2.1. Structural comparisons Structural comparisons of the ALDH crystal structures available in the PDB were made using PyMOL (http://www.pymol.org/) and Coot [10].

85

2.2. Sequence analyses

86

102

ALDH amino acid sequences were retrieved by Blastp searches at the UniProt site [11] (http://www.uniprot.org./blast). To identify protein sequences as members of the different reported ALDH families we carried out phylogenetic analyses using the MEGA6 software [12] (http://www.megasoftware.net/). Progressive multiple protein sequence alignments were performed with ClustalX version 2 [13] (http://www.clustal.org/). Analysis of residues at selected positions in the alignment was performed with ProfileGrid (http://www.profilegrid.org/). The unrooted phylogenetic tree was inferred from 500 replicates, using the Maximum Likelihood method. A discrete Gamma distribution was used to model evolutionary rate differences among sites (5 categories; +G, parameter = 1.0567). Sequence logos were constructed using the WebLogo server (http://weblogo.threeplusone.com). Each logo consists of stacks of amino acid letters, one stack for each position in the sequence. The height of letters within the stack indicates the relative frequency of each amino acid at that position [14].

103

3. Results and discussion

104

3.1. Structural analysis

105

114

As a first step to identify the structural factors relevant to the binding of the adenosine ribose of NAD+ and NADP+ and to analyze their different modes of binding, we examined the X-ray structures of ALDHs so far deposited in the PDB, both as apo forms and in complex with different nucleotides. The structures examined constitute a significant sample that includes 67 different ALDH enzymes from different organisms, belonging to 14 different already classified families, as well as eight structures of five enzymes that belong to families not yet classified by the ALDH Gene Nomenclature Committee.

115

3.2. Kinetic and equilibrium binding data analysis

116

In order to correlate the structural data with the coenzyme preference, we retrieved from the literature the kinetic and binding data of 74 enzymes, which belong to 13 classified families and to nine unclassified (Table 1). When analyzing the kinetic data we took into account several considerations. First, we assumed that the differences in the kinetic parameters values for the NAD+and NADP+-dependent reactions are mainly due to the absence or presence of the 20 -phosphate group in the adenosine ribose, given that, with the exception of ALDH3 enzymes, the structural analysis indicated that the rest of the molecule binds similarly, regardless of the nucleotide (not shown). Undoubtedly, the best parameter to assess binding is the dissociation constant (Kd) of the nucleotide from its complex with the enzyme, although the binding measured might not be kinetically relevant in some enzymes if non-productive binding occurs. For instance, a Kd(NAD+) of 10 lM was determined for ALDH1L1, but the enzyme exhibited a low activity

87 88 89 90 91 92 93 94 95 96 97 98 99 100 101

106 107 108 109 110 111 112 113

117 118 119 120 121 122 123 124 125 126 127 128 129 130 131

with NAD+ as coenzyme [15]. Values of Kd(coenzyme) are known for only a few ALDH enzymes, and in even fewer of them these values have been determined for both NAD+ and NADP+. Alternatively, Km values have been widely used as a measure of affinity. But in a Bi-Bi Ordered Steady-State mechanism, as the one followed by most ALDH enzymes, particularly the hydrolytic ones [9,16–18], Km is determined by the rate constant of several steps in the mechanism, not only by the on and off rate constants of the binding step. Indeed, in some ALDHs for which Kd and Km values are available there are important differences between them. The most striking reported differences between these two parameters are that of the above mentioned ALDH1L1 enzyme, which exhibits an apparent KmNAD+ of 2000 lM and a KdNAD+ of only 10 lM [15], and of the ALDH10 from the amaranth plant, which has a KmNAD+ of 39 lM and a KdNAD+ of 0.16 lM [18,19]. Also Kmnucleotide values may vary with the aldehyde substrate used. An example is the ALDH3I1 enzyme from Arabidopsis thaliana, which has a KmNADP+ of 1868 lM with hexanal as substrate but of 87 lM with nonenal [20]. Moreover, the interpretation of Km values is often complicated because apparent values obtained at a fixed aldehyde concentration, not always saturating, are frequently reported. True Kmnucleotide values, or even better Kia values, are very scarce in the literature of ALDH enzymes. To measure the affinity of the oxidized nucleotide kcat/Km should be determined, given that this kinetic parameter represents the bimolecular constant rate of the binding step of the first substrate to add to the enzyme in an ordered bi–bi mechanism, as the one generally found in ALDHs where the oxidized nucleotide binds first. But for several kinetically characterized enzymes kcat/Km is not known because Vmax or kcat values are not always reported even though Km values are. Besides, the low (kcat/Km)NAD(P)+ values for many ALDHs suggest that either there is an isomerization step associated with the binding of the coenzyme, or the kinetic mechanism is not ordered. In addition, the comparison of kinetic parameters of related enzymes is often complicated by the use of different assay conditions, mainly pH and temperature. In spite of these limitations, we tried to correlate the reported kinetic parameters with the known threedimensional structures and/or amino acid sequences of ALDHs in an attempt to recognize the residues that may influence their preference for the coenzyme, or their ability to bind both. From the data available in the literature, we calculated the ratios KmNAD+/ KmNADP+ and (kcat/Km)NAD+/(kcat/Km)NADP+, which were used as an indication of coenzyme preference.

132

3.3. Sequence analysis

175

A multiple sequence alignment of 1049 non-redundant ALDH protein sequences was performed to explore the degree of conservation of critical residues that influence coenzyme preference in the different ALDH families. In order to compare these residues among the different ALDH proteins, the sequence of the human ALDH2 mature protein (hALDH2) was used as reference. This analysis allowed us to infer the distribution of NAD+-, NADP+- or dual coenzyme specific enzymes within different ALDH families. Because some characterized ALDHs with a resolved crystal structure and/or known coenzyme preference do not belong to any of the previously reported ALDH families [21–23], the protein family assigned by the Conserved Domain Database [24] was given. We used this database because the protein classification included in it agrees with the previously reported ALDH families and with the results of the phylogenetic analysis performed in this work.

176

3.4. Structural determinants of coenzyme preference

191

In Fig. 1 are shown the interactions made by the adenosine ribose of NAD+ and NADP+ with the ALDH protein, as exemplified

192


133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174

177 178 179 180 181 182 183 184 185 186 187 188 189 190

193

CBI 7249


16 January 2015 3

L. González-Segura et al. / Chemico-Biological Interactions xxx (2015) xxx–xxx Table 1 Reported kinetic parameters for the nucleotides of different ALDH enzymes. Familya Organism Enzyme (UNIPROT/NCBI or PDB code) ALDH1/2 (cd07141) Homo sapiens ALDH1A1 (P00352) Homo sapiens ALDH1B1 (P30837) Ovis aries ALDH1A1 (1BXS) Rattus norvegicus ALDH1A2 (1BI9) Mus musculus ALDH1A3 (NP_444310.3) Homo sapiens hALDH2 (1O01, 1O05) Bos taurus ALDH2 (1A4Z, 1AG8) Rattus norvegicus ALDH2 (P11884) Drosophila melanogaster ALDH (Q9VLC5) Saccharomyces cerevisiae ALD6 (P54115) Saccharomyces cerevisiae ALD5 (P40047) ALDH1L1 (cd07140) Rattus norvegicus ALDH1L1 (2O2Q, 2O2R, 2O2P) ALDH3 (cd07087) Homo sapiens ALDH3A1 (3SZA) Rattus norvegicus ALDH3A1 (1AD3) Craterostigma plantagineum ALDH (Q8VXQ2) Arabidopsis thaliana ALDH3H1 (Q70DU8) Arabidopsis thaliana ALDH3I1 (Q8W033) ALDH4 (cd07083) Homo sapiens ALDH4A1 (3V9G) Thermus termophilus TTHA1578 (2BHP, 2EHQ) Bradyrhizobium japonicum PutA (3HAZ) ALDH5 (cd07103) Sus scrofa (NP_001231396)

Relevant residues (Predicted coenzyme preference)

NAD+

NADP+

KmNAD+/ KmNADP+

(kcat/Km)NAD+/(kcat/ Km)NADP+

Refs.

Km (Kd) (lM)

kcat/Kmb 1 1

Km (Kd) (lM)

kcat/Kmb 1 1

E195-Y224 (NAD+)

0.3 (0.7)

1.17 106

—c

—

E195-Y224 (NAD+)

3.6

3.38 105

NA

E195-Y224 (NAD+)

2.2

—

—

—

E195-Y224 (NAD+)

70

—

400

—

E207-F236

53 (0.2)

1.00 105

LA

E195-F224 (NAD+)

28 (49)

1.22 105

—

—

[44]

E195-F224 (NAD+)

47

—

—

—

[45]

E195-F224 (NAD+)

36

2.62 104

910

5.07 102

0.04

E214-F243 (NAD+)

63

—

1200

—

0.05

A199 (NADP+)

17,400

4.33 102

99

1.28 105

176

3.38 103

[48]

E216-S245 (NADP+)

6430

1.62

3470

1.23 102

1.9

1.32 102

[48]

Q600 (NADP+)

2000 (11)

—

1.5 (0.3)

3.30 104

1333 (37)

E140-G168 (NAD+)

21

5.52 104

990

1.29 103

0.02

43

[49]

E140-G168 (NAD+)

11

1.65 106

317

3.02 105

0.04

5.5

[5]

E143-A171 (NAD+)

459

1.92 104

9512

1.05 103

0.05

18

[50]

E149-A177 (NAD+)

421

4.37 104

NA

E212-G240 (NAD+)

71

1.99 105

1868

1.61 103

D236-D265 (NAD+)

100 (15)

1.00 105

—

—

E210-G240 (Dual)

71

2.25 104

120

3.50 103

E684-D713 (Dual)

420 (0.6)

8.10 103

—

—

[52]

E231-S260 (Dual)

310

—

—

—

[53]

(M

s

)

(M

s

)

[39]

[40]

[41]

0.18

[42]

[43]

52

[46]

[47]

[15]

[20]

0.04

124

[20]

[51]

0.59

6

[28]

(continued on next page)


CBI 7249


16 January 2015 4


Table 1 (continued) Familya Organism Enzyme (UNIPROT/NCBI or PDB code)


NAD+

Arabidopsis thaliana ALDH5F1 (Q9SAK4) Hordeum distichum L.d Solanum tuberosum (M1BFP7) Drosophila melanogaster (NM_143151) Escherichia coli GabD (3JZ4) Bacillus subtilis GabD (P94428)

E225-N254 (NAD+)

ALDH6 (cd07085) Rattus norvegicus ALDH6A1 (Q02253) Escherichia coli ECED1_0300 (B7MQA2) Pseudomonas aeruginosa MmsA (P28810) Bacillus subtilis IolA (1T90) ALDH7 (cd07130) Homo sapiens ALDH7A1 (2J6L) Acanthopagrus schlegelii (2JG7) Rattus norvegicus ALDH7A1 (Q64057) ALDH9 (cd07090) Homo sapiens ALDH9A1 (P49189) Sus scrofa (XP_005663205) Rattus norvegicus ALDH9A1 (Q9JLJ3) Pseudomonas aeruginosa PaBADH (2WME, 2WOX) Escherichia coli BADH (P17445) Xanthomonas translucensd ALDH10 (cd07110) Spinacia oleracea SoBADH (4A0 M) Amaranthus hypochondriacus AhBADH4 (O04895) Pisum sativum PsAMADH1 (3IWK) Pisum sativum PsAMADH2 (3IWJ) Solanum lycopersium SlAMADH1 (4I9B)

E192-N221 (NAD+) E204-N233 (NAD+) S182 (NADP+)

NADP+


Refs.

Km (Kd) (lM)

Km (Kd) (lM)

130

—

NA

[54]

166 31

— —

LA NA

[55] [56]

91

—

1200

—

44.6

7.73 104

(M

s

)

LA

kcat/Kmb 1 1

KmNAD+/ KmNADP+

kcat/Kmb 1 1

(M

s

)

0.08

[47] [57]

P160 (NAD+)

210

2.50 104

390

2.55 104

E212-Q241 (NAD+)

150

—

—

—

[59]

E212-D205 (Dual)

80

—

—

—

[60]

E176-G205 (Dual)

87

1.24 105

—

—

[61]

E179-A208 (Dual)

2300

4.78 102

—

—

[62]

P194 (NAD+)

454

3.05 102

—

—

[63]

P192 (NAD+) P221 (NAD+)

160

—

NA

200

—

—

—

[65]

P183 (NAD+)

14

1.12 105

—

—

[66]

P285 (NAD+) P183 (NAD+)

40

—

NA

28

—

1630

—

0.02

E179-S208 (Dual)

385 (85)

3.50 105

83 (45)

1.56 106

4.6 (1.9)

E179-V208 (Dual)

99

—

400

—

0.25

(Dual)

70

2.13 106

50

2.22 106

1.4

1

[70]

E185-L214 (NAD+)

20

9.95 104

320

4.64 103

0.06

21

[71]

E188-L217 (NAD+)

39 (0.16)

4.64 103

2500

—

0.02

E188-L217 (NAD+)

40

—

LA

[72]

E188-L217 (NAD+)

55

—

LA

[73]

E188-L217 (NAD+)

72

3.93 104

LA

[73]

0.54

1

[58]

[64]

[67] [68]

0.2

[9]

[69]

[18]


CBI 7249


16 January 2015 5

L. González-Segura et al. / Chemico-Biological Interactions xxx (2015) xxx–xxx Solanum lycopersium SlAMADH2 (B6ECN9) Zea mays ZmAMADH1 (4I8P) Zea mays ZmAMADH1b (G5DDC2) Zea mays ZmAMADH2 (C6KEM4) Avena sativad Pseudomonas sp. 13CM TMABADH-II (N0DT23) ALDH11 (cd07082) Streptococcus mutans SmGAPN (2EUH, 1EUH) Thermoproteus tenax TtGAPN (1KY8, 1UXT) Sulfolobus solfataricus GAPN-3 (Q97U30) ALDH14f (cd07093) Escherichia coli AldB (P37685) Rhodococcus erythropolis ThcA (P46369)

E188-L217 (NAD+)

89

—

LA

[73]

E190-L219 (NAD+)

91

2.61 104

LA

[73]

E191-L220 (NAD+)

79

—

LA

[73]

E190-L219 (NAD+)

86

—

LA

[73]

15 125

5.21 105 4.40 104

— NA

—

T180 (NADP+)

6000 (517)

2.50 103

24.5 (2.3)

2.73 106

I194 (Dual)

3300

9.97 103

NA

3100 21,100

— 6.74 101

20 mMe 90

— 1.98 105

65

2.89 104

E177-K206 (NAD+)

S201 (NADP+)

[74] [75]

245 (225)

9.16 104

[76]

[30]

234

3.40 104

[31] [77]

R197 (NADP+)

NA

E192-F220 (NAD+)

90

6.11 104

NA

[79]

E183-A212 (NAD+)

260

4.23 104

LA

[25]

E177-P206 (NAD+)

143

—

1000

—

0.14

ALDH26g (cd07092) Escherichia coli YdcW (1WNB, 1WND)

E175-R203 (NAD+)

54

1.05 105

484

4.37 103

0.11

24

[81]

Unnamed (cd07128) Burkholderia xenovorans BXE_A1420 (2VRO)

T185 (NADP+)

501

8.94 102

40

6.50 104

12.5

1.38 102

[82]

Unnamed (cd07129) Vibrio harveyi ALDH (1EYY)

T175 (NADP+)

382

1.48 105

1.6

5.73 106

239

2.58 102

[4]

390

1.56 105

1.4

6.07 106

279

2.57 102

[83]

439

—

9.58 104

ALDH25g (cd07119) Staphylococcus aureus SaBADH (4QN2, 4MPB) Bacillus subtilis GbsA (P71016)

Unnamed (cd07100) Synechococcus sp. PCC 7002 SYNPCC7002_A2771 (4ITA, 3VZ3) Salmonella typhimurium YneI (3EFV, 1EZ0) Gluconobacter oxydans 621H GOX0499 (3VZ0) Gluconobacter oxydans GOX1122 (Q5FRV6) Unnamed (cd07145) Sulfolobus tokodaii STK_00640 (Q976X5)

S167 (NADP+)

LA

P163 (NAD+)

43

8.84 105

480

P159 (NAD+)

160

7.35 104

NA

S160 (NADP+)

NA

I168 (Dual)

840

2.44 103

[78]

[80]

[84]

0.09

9

[85]

[86]

70

—

3810

9.74 102

[86]

0.22

3

[87]

(continued on next page)


CBI 7249


16 January 2015 6


Table 1 (continued) Familya Organism Enzyme (UNIPROT/NCBI or PDB code)


NAD+

Unnamed (cd07088) Escherichia coli AldA (2IMP, 2ILU) Unnamed (cd07113) Escherichia coli PAD (P80668)

NADP+

Km (Kd) (lM)

kcat/Kmb 1 1

Km (Kd) (lM)

E179-R208 (NAD+)

40

1.11 104

NA

E200-S229 (NAD+)

35

5.76 105

220

(M

s

)

kcat/Kmb 1 1 (M

s

KmNAD+/ KmNADP+


Refs.

)

[32]

4.09 104

0.16

14

[88]

NA, no activity was detected; LA, low activity was observed. a Protein families are according to Black and Vasiliou [21]; the accession numbers in parenthesis indicate the protein family assigned according to the Conserved Domain Database [24]. b Calculated from Vmax or kcat values, when reported. c No value for Km or kcat was reported. d The amino acid sequence of this enzyme has not been reported. e This enzyme exhibited positive cooperativity for binding of NADP+; therefore this is a S0.5 value. f ALDH14 family is defined according to Peng et al. [22]. g ALDH25 and ALDH26 are defined according to Riveros-Rosas et al. [23].

194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235

by those observed in the crystal structures of hALDH2 (PDB code 1O02; Fig. 1A) and succinic semialdehyde dehydrogenase from Synechococcus sp. PCC 7002 (unnamed ALDH family cd07100, PDB code 3VZ3; Fig. 1B), respectively. In most of the ALDH crystal structures in complex with NAD(H), the oxygen of the 20 -OH group of the adenosine ribose (O2B) is bound by strong hydrogen bonds to the side-chain amino group of Lys192 (hALDH2 numbering, which is the one used throughout this work) and carboxyl group of Glu195, in those enzymes that possess Glu in this position. In some of the latter enzymes, and in those with Gln195, there is also a weak hydrogen bond between the ribose O2B and the main-chain amide nitrogen of Gly225. The oxygen of the ribose 30 -OH group (O3B) makes strong hydrogen bonds with the main-chain carbonyl oxygen of residue 166 and with the side-chain amino group of Lys192. In the enzymes with NADP(H) bound, these two hydrogen bonds are generally conserved, as well as that of the amino group of Lys192 with O2B. The O1X of the 20 -phosphate group makes a hydrogen bond with the main-chain amide nitrogen of the residue at position 195, and in a few structures also with that at position 194, and/or with the side-chain groups of Ser or Thr at position 195 (when the enzyme has any of these two residues at this position) and Lys192. The O2X of the 20 -phosphate group is very often hydrogen bonded to the amino group of Lys192, and less often to the main-chain amide nitrogen of the residue at position 225, and to the carboxyl of Glu195. The O3X is making fewer interactions, accepting a hydrogen bond from the main-chain amide nitrogen of the residue at position 225, or from the guanidinium group of Arg225 in a few enzymes such as the ALDH11 from Thermoproteus tenax (PDB code 1KY8). In summary, the hydrogen bonds with the main-chain carbonyl oxygen at position 166, with the side-chain amino group of Lys192 and with the side-chain carboxyl group of Glu195—in those enzymes that have this residue—are the most important interactions for binding of the adenosine ribose of NAD+ or of NADP+, whereas the hydrogen bonds of the 20 -phosphate group with the main-chain amide nitrogen of the residue at position 195 and the side-chain amino group of Lys192 are the more general and important interactions for the binding of this group. The general architecture of the Rossmann fold of ALDH enzymes is very much conserved. Particularly, the loops b1-aA and b2-aB (Rossmann-fold nomenclature) have a very similar conformation in all examined crystal structures, as illustrated in Fig. 1C–E, so that

the main-chain atoms involved in the more important interactions with either the ribose or the ribose 20 -phosphate of the adenosine moiety are in almost identical positions in the ALDHs of known three-dimensional structure. The conformation of the loop b3-aC, which contains the residue 225, is more variable. Although in most enzymes this loop is close to the cavity where the coenzyme binds, so that the main-chain amide nitrogen of residue 225 may form weak hydrogen bonds with the O2B of the ribose (Fig. 1A) or with the O2X or O3X of the 20 -phosphate group, in other enzymes it is so distant that this interaction cannot be established (Fig. 1D). This may not be important for most of the enzymes that show this conformation of loop b3-aC because, as mentioned above, the interactions with the main-chain amide nitrogen of residue 225 are the weakest among those that the adenosine ribose of NAD(H) or NADP(H) may make with the protein. In some Glu195-containing enzymes the a-helix C is longer and the b3-aC loop shorter, which impede the binding of NADP+ (Fig. 1E). Therefore, the conformation of the polypeptide backbone in the region comprised between residues 221 and 227 may importantly affect the ability of the Glu195 enzymes to bind NADP+. In spite of this, in the majority of ALDH enzymes, the preference for NAD+ or NADP+ or the ability to bind both coenzymes is mainly determined by the nature and/or conformation of the side chain of residue 195. In addition, residues at positions 37, 53, and 224, although not directly involved in this binding, may also influence coenzyme specificity but only in the enzymes that have Glu195, as will be discussed below.

236

3.4.1. Residues at position 192 The results of multiple sequence alignments confirmed the previous conclusion—drawn in a study that used much fewer ALDH sequences [5]—that ALDH enzymes have a highly conserved Lys at this position, which is consistent with the many interactions that the Lys side chain makes with the adenosine ribose of both NAD+ and NADP+, as shown in Fig. 1A and B. In addition, in those enzymes that have Glu195, the positively charged amino group of Lys192 plays an important role in positioning the negatively charged carboxyl group of Glu195 for hydrogen bonding the adenosine ribose of NAD+, an interaction that seems to be important for the correct binding of this nucleotide. Not considering the ALDH20 sequences which have a Ser at position 192, only in 9 of the other 893 amino acid sequences analyzed Lys192 has been changed, and in all cases to a residue whose side-chain groups are hydrogen-

262


237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261

263 264 265 266 267 268 269 270 271 272 273 274 275 276

CBI 7249


16 January 2015 L. González-Segura et al. / Chemico-Biological Interactions xxx (2015) xxx–xxx

7

Fig. 1. Binding site of the adenosine ribose in ALDH enzymes. (A) Interactions of the NAD+-adenosine ribose as observed in the crystal structure of hALDH2 (ALDH1/2; PDB code 1O01). This mode of binding is representative of most of the NAD+-specific ALDHs. (B) Interactions of the NADP+-adenosine ribose 20 -phosphate as observed in the crystal structure of succinic semialdehyde dehydrogenase from Synechococcus sp. PCC 7002 (unnamed ALDH family cd07100, PDB code 3VZ3), which exemplifies the binding of this nucleotide observed in the crystal structure of enzymes with Ser/Thr. Numbering of the residues according to that in hALDH2 is given in parenthesis. Hydrogen bonds (cut off 3.5 A) are depicted as black dashed lines. (C) Superposition of the loops that contain residues involved in binding the adenosine ribose in hALDH2 (code 1O01; green) with those of PaBADH (ALDH9, PDB code 2WME; cyan). The same conformation of these loops was observed in the reported crystal structures of ALDH1/2, ALDH1L1, ALDH3, ALDH4, ALDH5, ALDH6, ALDH7, ALDH9, ALDH10, ALDH11, and ALDH26 enzymes, as well as in those of the unnamed families cd07139 and cd07088. (D) As in (C) but comparing the loops of hALDH2 (PDB code 1O01; green) with those of the BXE_A1420 from Burkholderia xenovorans LB400 (unnamed ALDH family cd07128, PDB code 2VRO; cyan), which is representative of enzymes of the ALDH20 (PDB code 3K9D) and ALDH25 (PDB code 4MPB) families, as well as of GOX0499 from G. oxydans 621H (unnamed ALDH family cd07100, PDB code 3VZ0), GabD1from L. acidophilus (unnamed ALDH family cd7100, PDB code 3ROS), YNE1 from Salmonella typhimurium (unnamed ALDH family cd07100, PDB code 3EFV) and NADP+-specific ALDH from V. harveyi (unnamed ALDH family cd07129, PDB code 1EYY). (E) As in (C) but comparing the loops of hALDH2 (PDB code 1O01; green) with those of observed only in the succinic semialdehyde dehydrogenase from Homo sapiens (ALDH5, PDB code 2W8R; cyan). (F) Superposition of different residues at position 192, as observed in hALDH2 in complex with NAD+ (ALDH1/2, PDB code 1O01; green), succinic semialdehyde dehydrogenase from Brucella melitensis (ALDH5, PDB code 3EK1; magenta) and the non-phosphorylating glyceraldehyde-3-phosphate dehydrogenase from Methanocaldococcus jannaschii (unnamed ALDH family cd07094, PDB code 3PQA; yellow). The portion of the NAD+ molecule shown (black carbons) is that observed in the hALDH2 crystal. (G) Superposition of the three observed conformations of Glu195: (1) ‘‘NAD+-compatible’’, as observed in hALDH2 in complex with NAD+ (ALDH1/2, PDB code 1O01; green); (2) ‘‘NADP+-compatible’’ in E. coli lactaldehyde dehydrogenase in complex with NADPH (unnamed ALDH family cd07088, PDB code 2ILU; cyan); and (3) ‘‘NADP+-compatible’’ in PaBADH in complex with NADP+ (ALDH9, PDB code 2WME; yellow). Only the side-chain of Glu195 is shown for the two latter enzymes, as the rest of the protein superpose to that of hALDH2. The portion of the NADP+ molecule shown (black carbons) is that observed in the PaBADH crystal. Amino acid side-chains or main chain atoms are depicted as sticks with carbon atoms colored according to the protein shown, oxygen in red and nitrogen in blue. The figure was generated with PyMOL (http://www.pymol.org/). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) 277 278 279 280 281 282 283

bond donors (His, Gln, Asn, Arg or Ser) and have conformations similar to that of Lys192 (Fig. 1F). These residues may therefore participate in the binding of the adenosine ribose similarly to Lys192. The neutral Ser, Gln or Asn residues, however, cannot exert the electrostatic effect that Lys192 has on the side chain of Glu195 favoring its appropriate conformation for NAD+ binding, and this may explain why Gln/Asn192 are present in the NADP+-specific

enzymes that at position 195 have Ser or Ala that cannot make hydrogen bonds with the hydroxyl groups of the bound ribose because of their shorter side-chains. Six of the nine above mentioned exceptions belong to the ethanol inducible ALDH14s, a protein family described first by Peng et al. [22]. In summary, from the crystallographic data it is clear that the residue at position 192 plays a pivotal role in the binding of NAD(H) and NADP(H), but it


284 285 286 287 288 289 290

CBI 7249


16 January 2015 8 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354


is also clear that it does not confer coenzyme specificity, as previously discussed and proved by others [5]. 3.4.2. Residues at position 195 In this work we confirmed and extended the previous observation [5] that the main determinant of nucleotide specificity in ALDH enzymes is the side chain of the residue at position equivalent to 195, which most frequently is Glu. But our sequence analysis indicate that can also be Asp, Gln, Arg, Lys, His, Pro, Thr, Ser, Ile, Leu, Val, Ala, and Gly. Crystal structures of enzymes having Glu, Asp, Gln, His, Pro, Thr, Ser, Ile and Ala have been reported but not yet one that has Arg or Lys this position. It might be interesting that no enzyme with any of the aromatic residues, Phe, Tyr or Trp, or sulfur-containing residues, Cys or Met, at position 195 has been so far reported. On the basis of the known crystal structures and of the published kinetic data, below we analyze the effect that the nature and conformation of residue at position 195 has on coenzyme preference. Q3 3.4.2.1. ALDH enzymes with Glu195. As mentioned above, the reported crystal structures of ALDH enzymes that have Glu at position 195 show that its side-chain carboxyl group is hydrogen bonded to the 20 -OH, and less often to the 30 -OH, of the ribose of the adenosine moiety of NAD(H), thus contributing to their binding. Most of the biochemically characterized enzymes having this residue prefer NAD+ to NADP+ (Table 1) because the Glu195 side-chain, in the conformation generally observed in apo and NAD(H)-bound enzymes, sterically and electrostatically interferes with the binding of the 20 -phosphate group of NADP+. The possibility of binding NADP(H) by these enzymes depends on the conformational flexibility of Glu195, i.e., on whether its side chain can or cannot move from the position that participates in NAD(H) binding to other positions that leave enough room to accommodate the 20 -phosphate group of NADP(H). A comparison between the three conformations of the Glu195 side-chain so far observed in crystal structures of ALDHs is shown in Fig. 1G; the conformation marked (1) is incompatible with the binding of NADP+ while those marked (2) and (3) are compatible with NADP+ binding. The conformational flexibility of Glu195 is allowed or restricted by nearby residues, particularly the one at position 224 and those at positions 37 and/or 53 that in some cases determine the conformation of the 224 residues. The reported crystal structures provide examples of the effect of these residues on coenzyme preference, as described below. 3.4.2.1.1. Glu195–Phe/Tyr/Leu224. The bulky aromatic side-chain of Phe or Tyr at position 224, as in hALDH2 (PDB code 1O01; Fig. 2A), hinders the movement of Glu195. Although the side chain of the aromatic residue is exposed to the solvent, it cannot adopt a conformation that would not obstruct the movement of Glu195 sidechain because the aromatic ring is held in the position observed in the crystals by l-stacking interactions with a Phe at position 37. In plant ALDH10s (PDB codes 4A0M, 3IWJ, 3IWK, 4I8P, and 4I9B), a Leu224 stabilized by van der Waals contacts with Ile/ Leu37 (Fig. 2B) also impedes the movement of Glu195 and, therefore, the binding of NADP+. Multiple sequence alignments indicate that the residues Glu195, Phe/Tyr/Leu224 and Phe/Tyr/Ile/Leu37 are well conserved in the enzymes of the ALDH1/2 (88%), ALDH10 (85%), ALDH27 (66%) and ALDH14 (47%) families, and in some bacterial ALDH9 and ALDH27 enzymes. 3.4.2.1.2. Glu195–Arg224. Another group of ALDH enzymes that possess Glu195 is formed by those that have Arg at position 224, whose guanidinium group is exposed to the solvent and held in this position by hydrogen/ionic bonds with a Glu or Asp at position 53 and/or with Gln or Glu at position 226. The crystal structure in complex with NADPH of the lactaldehyde dehydrogenase from Escherichia coli (unnamed ALDH family cd07088, PDB code 2ILU)

shows that Glu195 can move to a conformation compatible with NADP(H) binding. In this conformation, the side-chain carboxyl group of Glu195 is exposed to the solvent and hydrogen bonded to the 20 -phosphate group, which suggests that one of two groups are protonated. In the E. coli lactaldehyde dehydrogenase, the carboxyl cannot move farther away from the 20 -phosphate because it would crash with the side chain of Arg224 (Fig. 2C). Therefore, even though there is no steric hindrance for binding of the 20 -phosphate group, the electrostatic repulsion with the carboxyl has to be cancelled out, which has an energetic cost that is probably the reason behind the preference for NAD+ of some of these enzymes (Table 1) as the aminoaldehyde dehydrogenase YcdW from E. coli (ALDH26). But this does not explain the reported lack of activity with NADP+ of the E. coli lactaldehyde dehydrogenase (Table 1). Multiple sequence alignments indicate that many ALDH26s (44%), some bacterial ALDH5s and ALDH9s, as well as the known ALDH10s from chlorophyta, have Glu195 and Arg224 (Fig. 3). We predict that these enzymes would prefer NAD+ but they could also bind NADP+, with the exception of the few ALDH5s that may have the abnormal helix aC observed in the human ALDH5 (PDB code 2W8R; Figs. 1E and 2D). Their low affinity for NADP+, however, would not warrant their classification as dual coenzyme-specificity enzymes. We cannot validate this prediction, however, as most of these enzymes have not been yet kinetically characterized. 3.4.2.1.3. Glu195–Asp224. A significant proportion of bacterial ALDH9s and some bacterial ALDH4s, ALDH5s, ALDH6s, ALDH26s and of the unnamed family cd07138 have Glu195 and Asp at position 224. The reported crystal structures of some of them (ALDH5, PDB code 3IFG; ALDH6, PDB code 4E4G; ALDH9, PDB code 3R31; ALDH4, PDB code 3HAZ; unnamed ALDH family cd07138, PDB codes 3I44 and 3TY7) show that the Asp224 side-chain is directed toward the solvent away from Glu195 (Fig. 2E), most likely because of electrostatic repulsion, and that the loop b3-aC is more distant from the loop b2-aB than in the rest of the Glu195 containing ALDHs of known three-dimensional structure. Therefore, there is enough room for the Glu195 carboxyl group to move sufficiently far away from the 20 -phosphate group of NADP+ to avoid electrostatic repulsion. A small or positively charged residue at position 53 would favor this NADP+-compatible position of Glu195. We predict that these enzymes could bind both NAD+ and NADP+, i.e., have dual-coenzyme specificity. We cannot test out this prediction with data in the literature because, to the best of our knowledge, the two enzymes of this group that have been biochemically characterized were not assayed with NADP+ (Table 1). 3.4.2.1.4. Glu195–Ser/Thr/Ala/Gly224. As in the rest of ALDHs with Glu195, these enzymes can bind NAD+. But the presence of a small neutral residue at position 224 would allow the side chain of Glu195 to move and adopt a NADP+-compatible conformation in which the carboxyl group may be close to or distant from the 20 phosphate group depending on the size and/or polarity of the residue at position 53. Therefore, they could bind NADP+ with lower affinity than NAD+, or have true dual-coenzyme specificity. In some enzymes, such as the BADH from Staphylococcus aureus (ALDH25, PDB code 4QN2; Fig. 2F), an Arg at position 37, bonded to a Glu at position 53, could interact with the carboxyl of Glu195 in an NADP+ compatible conformation in which the carboxyl is close to the 20 -phosphate group. Because of this, we predict that the affinity for NADP+ would be lower than that for NAD+, for the reasons given above, and that the affinity for the latter would be relatively low compared with other Glu195 enzymes, because this residue would be not fixed in the ideal NAD+-compatible conformation. Indeed, the report of its kinetic characterization indicated a low affinity for NAD+, and also that it can use both NAD+ and NADP+ although the activity with NAD+ is ten-times higher than with NADP+ [25]. This is probably true for most ALDH25 enzymes because 91% of them posses Arg37 and Glu53. In contrast, in the case of the phe-


355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420

CBI 7249



9

Fig. 2. Comparison of the adenosine-ribose binding-pocket in different ALDH enzymes. Surface representation showing the relevant residues as sticks with carbon atoms colored green, nitrogen blue, and oxygen red. Hydrogen bonds (cut off 3.5 A) are depicted as black dashed lines. Numbering of the residues according to that in hALDH2 is given in parenthesis. The figure was generated with PyMOL (http://www.pymol.org/). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441

nylacetaldehyde dehydrogenase (PAD) from E. coli (unnamed ALDH family cd07113, UNIPROT code P80668) that has Glu195 and Ser224, we predict that the presence of a Leu37 and Asp53 would disfavor the movement of Glu195 towards a NADP+-compatible conformation, which is in agreement with the results of its kinetic characterization that showed a relatively high affinity for NAD+, suggested by the Km value, and a ten-times higher preference for NAD+ over NADP+ as indicated by the ratio of the kcat/Km for both coenzymes (Table 1). As just mentioned, the kinetic data for most of the enzymes of this group that have been characterized confirm that they prefer NAD+, but one of them, the ALDH9 from Pseudomonas aeruginosa (PaBADH), binds NADP+ with higher affinity than NAD+ [9] (Table 1). The crystal structure of this enzyme in complex with NADP+ (PDB code 2WME; Fig. 2G) showed that the side chain of Glu195 has moved away from the 20 -phosphate group and forms an ionic bond with an Arg at position 53 [26]. This eliminates the electrostatic repulsion between the carboxyl and 20 -phosphate groups, thus facilitating NADP+ binding. Although in this conformation the carboxyl group of Glu195 does not participate in binding the hydroxyl groups of the adenosine ribose, the 20 -phosphate

group makes six additional hydrogen bonds with the protein, which accounts for the better binding of NADP+ than NAD+ by this enzyme. Glu195 would be portioned between two conformations, one compatible with binding NADP+, stabilized by Arg53, and the other compatible with binding of NAD+, stabilized by Lys192. The possibility of adopting these two conformations possibly explains the relatively low affinity for NADP+ of PaBADH when compared with NADP+-specific enzymes, and its relatively low affinity for NAD+ compared with NAD+-specific Glu195-containing enzymes. Since the differences in affinity for the two coenzymes are not large [9], PaBADH can be considered as having dual-coenzyme specificity. Sequence analysis indicated that the other ALDH9 enzymes from the genus Pseudomonas also have Ser224 and Arg53 (Fig. 3). In the ALDH1/2 enzyme from yeast that have been named ALD5, and in the ALDH6 of some non-mammalian animals, a Lys at position 53 may fulfill the same role than Arg53 in PaBADH, keeping the carboxyl group of Glu195 away from the 20 -phosphate group in the NADP(H) complexes. The ALDH1/2 enzyme for yeast exhibited an abnormally high Km for both nucleotides and abnormally low kcat/Km values, showing that both nucleotides, particularly NAD+, are poor substrates for unknown reasons. Both Km values


442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462

CBI 7249


16 January 2015 10 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528


only differ by a factor of 1.9, but the catalytic efficiency for NADP+ is about 80-times higher than that for NAD+, indicating that this enzyme is has a clear preference for NADP+ (Table 1). There are no kinetic data available for the ALDH6 enzymes. Special cases of ALDH enzymes that have Glu195 and a small residue at position 224—a glycine, or an alanine in a few cases— are most of the ALDH3s of known sequence. In most of these enzymes the side chain of Glu195 can freely move away from the 20 -phosphate group of NADP+ because they have a truncated amino-terminal and lack residues at position 37 and 53. It could be predicted therefore that these enzymes could bind both NAD+ and NADP+ with similar affinity, i.e., have dual-coenzyme specificity. The published ALDH3 structures in complex with NAD+ (ALDH3 from rat, PDB code 1AD3; Fig. 2H) or NADP+ (ALDH3 from Pseudomonas putida, PDB code 3LV1) show that indeed they can bind both nucleotides. Moreover, in the 3LV1 structure, the side chain of Glu195 adopts a conformation similar to the NADP+-compatible conformation observed in PaBADH, i.e., far away of 20 -phosphate group and making a hydrogen bond with Tyr53. It has been reported, however that this enzyme exhibits greater affinity for NAD+ than for NADP+—although the kinetic parameters have not been given [27]—, as do most of the rest of ALDH3 enzymes that have been biochemically characterized (Table 1). The reason for their NAD+ preference appears not to be steric or electrostatic impediments that discriminate against binding of the 20 -phosphate group of NADP+, as in other NAD+-specific enzymes, but most likely the differences in the conformation of the whole adenosine moiety between NAD+ and NADP+ when bound to the enzyme, as seen in the crystal structures of the rat and Pseudomonas ALDH3 enzymes, which may result in fewer interactions of the NADP+ with the protein. Furthermore, in these two enzymes the adenosine moiety of the nucleotide has a different position than the one generally observed in most of the rest of the ALDHs of known structure. This is probably due to the altered conformation of the adenosine-binding pocket of ALDH3 proteins because of their truncated N-terminal region. 3.4.2.1.5. Glu195–Val224. A valine at position 224 allows the movement of the Glu195 side chain to make room for NADP+ binding, but only up to a position where it makes a hydrogen bond with the 20 -phosphate group, as shown in the crystal structure of pyrroline-5-carboxylate dehydrogenase from Thermus thermophilus (ALDH4, PDB code 2BHP). The energetic cost of offsetting the electrostatic repulsion between these two groups may be the reason behind the preference of NAD+ over NADP+ of these enzymes [28]. Crystal structures of enzymes with Glu195 and Val224, besides the ALDH4 from T. termophilus discussed above, are scarce: the E. coli BADH (ALDH9; UNIPROT code P17445), the Dictyostelium discoideum ALDH2 (UNIPROT code Q54FY) and the Anabaena sp. ALDH6 (UNIPROT code BAB75470). Of these, to the best of our knowledge, only the E. coli BADH has been kinetically characterized. In agreement with our predictions, it can bind both nucleotides and exhibits a 4-times higher affinity for NAD+ than for NADP+, as indicated by Km values, which justifies classifying it as a dual-specificity enzyme (Table 1). In summary, all Glu195-containing ALDHs bind NAD+ with high affinity because the interactions that the carboxyl makes with the adenosine ribose, but some of them can also bind NADP+ because of the three different conformations that this residue may adopt: one that participates in NAD+ binding, observed in apo as well as in NAD+-bound enzymes (Fig. 1A), and the two that are compatible with binding of NADP+ observed in enzymes with NADP(H) bound (Fig. 1B and G). The possibility that Glu195 adopts any of the two ‘‘NADP+-compatible’’ conformations depends on the size or conformation of the residue at position 224, as discussed. In one of the two NADP+-compatible conformations the carboxyl group is close and hydrogen-bonded to the 20 -phosphate group, whereas in the

other, energetically more favorable and that results in higher affinity for NADP+, the carboxyl is distant from the 20 -phosphate group because an interaction by hydrogen/ionic bonds with a polar residue at position 53. On the basis of our analysis, we propose that genuine dual-coenzyme specificity ALDH enzymes would be those that have Glu195, a small neutral residue at position 224 and a residue at position 53 that by interacting with the Glu195 carboxyl group holds it away from the 20 -phosphate, thus preventing electrostatic repulsion between the two negatively-charged groups.

529

3.4.2.2. ALDH enzymes with Asp195. Multiple amino acid-sequence alignments showed that many animal ALDH4 and a significant proportion of ALDH26 enzymes have Asp195 (Fig. 3), whose side chain is directed toward the solvent, as observed in the crystal structures of apo and NAD+-complexed ALDH4 enzymes from mouse and human (PDB codes 3V9L and 3V9G, respectively; Fig. 2I). The ALDH4 enzymes frequently have a Tyr at position 53 that stabilizes the Asp195 conformation by a hydrogen bond with the carboxyl group. Most of these enzymes have an Asp at position 224, but this residue apparently does not have any effect on the Asp195 conformation because the two carboxyl groups are very far apart. In contrast to Glu195, the shorter side-chain of Asp195 does not interact with the hydroxyl groups of the adenosine ribose, and therefore does not contribute to the binding of NAD+. Neither would the Asp195 carboxyl sterically impede the binding of NADP+ but, as the 20 -phosphate group would be placed at a short distance from the carboxyl group, electrostatic repulsion between them would occur, resulting in a low affinity for NADP+. Therefore, we predict that the ALDH4 enzymes having Asp195 would prefer NAD+. Yet, the affinity for NAD+ would be lower than that of the Glu195 enzymes because the NAD+-bound structure shows that not only the hydrogen bonds between the adenosine ribose and carboxyl group of Asp195 are absent, but also those between the ribose and the main-chain amide groups of residues 194 and 225. The KmNAD+ and KdNAD+ values of the human ALDH4 enzyme (Table 1) support our predictions. NADP+ has not been tested yet. The ALDH26 enzymes have a nonpolar residue, Val/Ile/Leu, at position 53. Since these residues cannot make ionic or hydrogen bonds with Asp195, it is possible that the conformation of the latter residue is not the same as that observed in the ALDH4s. Even though, it is to be expected that their nucleotide preference and their affinity for NAD+ will be similar to that of the ALDH4s. None of the ALDH26 with Asp195 has been kinetically characterized so far.

538

3.4.2.3. ALDH enzymes with Gln195. Among the ALDH enzymes used in our multiple sequence alignments, a Gln at position 195 is present in almost all known ALDH1L1 enzymes (Fig. 3), and also in some ALDH4s (e.g., S. cerevisiae; UNIPROT code P07275), ALDH9s (e.g., Ciona intestinalis; AK112741), and enzymes of the unnamed ALDH family cd07088 (e.g., Pseudoalteromonas atlantica T6c; PDB 3K2W). The only three-dimensional structures known in complex with the coenzyme are those of the rat ALDH1L1 with NADP+ or NADPH bound (PDB codes 2O2Q and 2O2R, respectively). The side chain of Gln195 is directed towards the solvent, away from the ribose-binding cavity, and not obstructing the binding of the 20 -phosphate group, neither making any interaction with this group, which is mainly bound by hydrogen bonds to the side-chain amino group of Lys192 and to the main-chain amide nitrogens of Gln195 and Gly225 (Fig. 2J). In the crystal structure of the apo form of enzyme from P. atlantica, the side-chain amide group of Gln195 is also directed towards the solvent in a position that could not make hydrogen bonds with the ribose hydroxyls groups of NAD+. This conformation of the Gln195 side-chain, opposite to that of the Glu195 side-chain in a NAD+-compatible conformation but similar to the one of this side chain compatible with NADP+ binding observed in PaBADH, is probably the consequence of the

571


530 531 532 533 534 535 536 537

539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570

572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592

CBI 7249



11

Fig. 3. Phylogenetic analysis and sequence logos of selected positions in different ALDH families. An unrooted phylogenetic tree that includes 1049 ALDH protein sequences is shown; sequences belonging to the same family are indicated with different colors. The tree was inferred from 500 replicates, using the Maximum Likelihood method. The best tree with the highest log likelihood (367547.8222) is shown. The branches in the unrooted tree are drawn to scale; the bar length indicates the number of substitutions per site. Protein families are identified according to the ALDH Gene Nomenclature Committee classification [21]; ALDH14 is defined according to Peng et al. [22]; ALDH25, ALDH26 and ALDH27 are defined according to Riveros-Rosas et al. [23]. Logos showing conservation of residues 195 and 224 (according to the sequence of mature hALDH2) are shown adjacent to each ALDH family included in the tree. The amino acids color scheme of the logos is according to their chemical properties: polar (G, S, T, C, Q, N), green; aromatic (Y, W, F), purple; basic (K, R, H), blue; acidic (D, E), red; and hydrophobic (A, V, L, I, P, M), black. Possible coenzyme preference of ALDH enzymes is indicated at the right side of each logo. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614

absence of an electrostatic effect of Lys192 on the neutral sidechain of glutamine. ALDH1L1 has been biochemically characterized and found that it has a clear preference for NADP+ over NAD+, particularly when the Km values for both nucleotides are compared (Table 1). But although the apparent KmNAD+ is 2000 lM, a KdNAD+ value of only 11 lM [15] indicates that the enzyme has a relatively good affinity for this coenzyme, although lower than that of the Glu195 enzymes. Other factors should be involved in determining the high KmNAD+ value. One possibility is that the hydrogen bond of the 20 -OH of the adenosine ribose with the side-chain of Glu195 is very important for the correct binding of the coenzyme, and that in the absence of this interaction, NAD+ binds in a kinetically ‘‘lowproductive’’ manner. This could also account for the low activity with NAD+ of other ALDHs that instead of Glu195 have a residue with a side-chain that cannot hydrogen-bond the adenosine ribose. 3.4.2.4. ALDH enzymes with Pro195. Sequence alignments showed that most of the animal ALDH9 and of the eukaryotic ALDH7 enzymes have Pro at position 195. The known crystal structures with NAD+ (ALDH7, PDB code 2JG7; ALDH9, PDB code 1BPW; and unnamed ALDH family cd07100, PDB code 3EFV) or NADH bound (ALDH7, PDB code 2J6L) indicate that the side chain of Pro195 leaves no room for binding the 20 -phosphate group of the adeno-

sine ribose of NADP+ (Fig. 2K). Given the conformational rigidity of this residue, it could not adopt a NADP+-compatible conformation, which is consistent with the finding that these enzymes clearly prefer NAD+ to NADP+ (Table 1). But there is one exception, that of the ALDH5 from Bacillus subtilis (UNIPROT code P94428), which showed dual-specificity (Table 1). The structural bases for this abnormal behavior, when compared with the rest of the enzymes that have Pro195, are not known since there is no crystal structure of this enzyme.

615

3.4.2.5. ALDH enzymes with Thr/Ser195. By crystallographic and mutational studies of the NADP+-dependent ALDH from Vibrio harveyi (unnamed ALDH family cd07129) it was proven long ago that a Thr at position 195 confers a great affinity for NADP+ to this enzyme [4,29]. Its crystal structure in complex with NADP+ (PBD code 1EYY) shows the side-chain hydroxyl and the main-chain amide nitrogen making hydrogen bonds with the 20 -phosphate of the nucleotide. Thr/Ser195 makes the same interactions in other crystal structures (ALDH11, PDB code 2EUH; unnamed ALDH family cd07100, PDB code 3VZ3; unnamed ALDH family cd07150, PDB code 4H73), as shown in Fig. 1B. The Thr/Ser195 residue has the same conformation in the apo forms of these enzymes (ALDH11, PDB code 3PRL; unnamed ALDH family cd07129, PDB code 3V4C;

624


616 617 618 619 620 621 622 623

625 626 627 628 629 630 631 632 633 634 635 636

CBI 7249


16 January 2015 12 637 638 639 640 641 642 643 644 645 646 647

648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670

671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699


unnamed ALDH family cd07094, PDB code 3PQA; and unnamed ALDH family cd07151, PDB code 3R64), which indicates that the arrangement of this residue is optimal for binding of NAD(P)H even in the absence of the nucleotide. It is not however optimal for binding of NAD+, whose ribose could not make any interaction with the short side-chains of Thr/Ser195. Given the fewer interactions that the ribose can make, NAD+ could bind but with lower affinity than NADP+, as it has been experimentally proven with some of these enzymes (Table 1). Our sequence analysis showed that enzymes with Thr/Ser195 are almost all ALDH11s, and some bacterial ALDH4, ALDH5 and ALDH14 enzymes (Fig. 3). 3.4.2.6. ALDH enzymes with His/Arg/Lys195. In the only reported crystal structure of an enzyme with His195, that of the GabD1 from Lactobacillus acidophilus (unnamed ALDH family cd07100, PDB 3ROS; not shown), the His side-chain is directed towards the solvent and makes a hydrogen bond with a sulphate anion that occupies a position similar to that of the 20 -phosphate group of NADP+ in other enzymes. Although the kinetics of this enzyme have not been studied, its crystal structure suggests that it is specific for NADP+. Regarding those enzymes that have Arg195, there is no crystal structure yet but one of them, the aldB from E. coli (ALDH14, UNIPROT code P37685) has been kinetically characterized and found to be NADP+-specific, which suggests that the Arg side chain is directed towards the solvent, not only not opposing the binding of NADP+ but probably favoring it by interactions with the 20 -phosphate group. Of the enzymes with Lys195 nothing is known apart from the amino acid sequence of a few of them. Sequence analysis showed that His195 is also present in ALDH from Ferroplasma acidarmanus (unnamed ALDH family cd07078, UNIPROT code S0ATQ1), in addition to the above-mentioned GabD1 enzyme. Arg195 is present also in the ALDH26 from Nocardia farcinica IFM 10152 (UNIPROT code Q5Z0C3), and an ALDH3 from Dictyostelium and Entomoeba genuses. Lys195 is present in ALDH3 from Entamoeba histolytica (UNIPROT code P30840). 3.4.2.7. ALDH enzymes with Ile/Leu/Val195. The crystal structure of the ALDH11 from T. tenax in complex with NAD+ (PDB code 1UXT) shows that the main interactions of the protein with the adenosine ribose of NAD+ are the hydrogen bonds with the amino group of Lys192, the main-chain carbonyl oxygen of the residue at position 166, and the amide nitrogen of Ser194. In this crystal structure and in that of the ALDH4 enzyme from Mycobacterium tuberculosis (PDB code 4IHI), the side chain of Ile195 is distant from the ribose, so it leaves enough room for the 20 -phosphate of NADP+ to bind. In fact, Ile195 has the same conformation in the structure of the T. tenax enzyme with NADP+ bound (PDB code 1KY8; Fig. 2L). The 20 -phosphate group of NADP+ makes additional hydrogen bonds with the main-chain amide nitrogen of Ile195 and Gly225 as well as with the hydroxyl group of Ser194. On the basis of these crystallographic data, poor binding of NAD+ could be expected, which is consistent with the high KmNAD+ of this enzyme (Table 1), and a better binding of NADP+. However, NADP+ was reported not to be a substrate, but rather an inhibitor of the NAD+-dependent reaction [30], indicating that it does bind but in a non-productive manner. Later, it was found that NADP+ does act as coenzyme but exhibited positive cooperativity and a very high S0.5, around 20 mM [31] (Table 1). The reason for this abnormal behavior is not apparent to us from the crystallographic data of the ribosebinding region of the NADP+ complex. Among the ALDH amino acid sequences analyzed, the enzymes with Ile/Val195 are very scarce and they belong to the ALDH11 and ALDH4 families. Enzymes with Leu195 are in the ALDH7 and ALDH5 families, but they have not been biochemically or structurally characterized.

3.4.2.8. ALDH enzymes with Gly/Ala195. Sequence alignments indicated that a few bacterial enzymes of the ALDH14 and ALDH5 families as well as ALDH2 enzymes from yeasts (as the one from Saccharomyces cerevisiae UNIPROT code P54115) have a small apolar residue at position 195 and Ala at position 194 (Fig. 3). The crystal structure of a M. tuberculosis enzyme (unnamed ALDH family cd07139, PDB code 3B4W) shows the adenosine ribose of NAD+ making hydrogen bonds with only the amino group of Lys192 and the main-chain of the carbonyl oxygen of the residue at position 166. The small side-chain of Ala195 would not interfere with binding of the 20 -phosphate group, which probably would make additionally interactions with the protein. Therefore, these enzymes most likely bind NADP+ better than NAD+, given the few interactions that the protein would make with the ribose. To date, none of these enzymes has been kinetically characterized.

700

3.5. Mutant ALDH enzymes with altered adenosine ribose binding-site

715

The first mutagenesis study that resulted in changes in coenzyme affinity was performed in the NADP+-specific ALDH from Vibrio harveyi (unnamed ALDH family cd07129) where two Gly residues at positions 245 and 250 were mutated to Ala [29]. The mutant enzyme Gly245Ala totally lost its ability to bind either NAD+ or NADP+, which is probably due to the narrowing of the tunnel through which the nicotinamide enters into the active site. The mutant Gly250Ala exhibited decreased kcat/Km values for both nucleotides, particularly in the NAD+-dependent reaction. Then, this research group replaced the Thr at position 195 of the same enzyme to Ser, Gln, Asn, Glu and Asp [4]. The two latter mutations changed the specificity to NAD+, due to increases in the affinity for this nucleotide together with decreases in the affinity for NADP+. The change to a non-charged polar residue increased the affinity for NAD+, without affecting that of NADP+. Our analysis of the role of these residues, above, is in full agreement with the results of this study. Perozich et al. [5], changed Lys192 for Gln demonstrating the pivotal role of this residue in NAD(P)+ binding. They also mutated Glu195 and found that whereas the catalytic efficiency in the NADP+-dependent reaction progressively increased in the mutants Glu195Asp, Glu195Asn and Glu195Gln, that in the NAD+-dependent reaction decreased in the same order, in total accordance with the available crystallographic data discussed above. In an attempt to increase the affinity for NADP+ of the lactaldehyde dehydrogenase from E. coli, Phe196 was substituted with Thr [32]. Our sequence analysis showed that this is a highly variable position in ALDH enzymes in general, although it shows a certain degree of conservation within particular families. For instance, Thr is present in most ALDH7s, Gln in most ALDH1/2s and ALDH11s, a hydrophobic bulky residue in ALDH3s, ALDH10s, ALDH25s, and in many ALDH9s, Arg or Lys in most ALDH6s, and Glu, Asp, Gln, Ser, Leu or Phe in ALDH5s. The mutant Phe180Thr exhibited increased affinity for both NAD+ and NADP+, and a change of the rate-limiting step from release of the reduced nucleotide in the wild type enzyme to deacylation in the mutant. The crystal structures of this enzyme with NADH and NADPH (PDB codes 2IMP and 2ILU, respectively; Fig. 2C) do not show any direct relation of this residue with Glu195 in either of its two conformations, i.e., in the NAD+- or NADP+-compatible. But it is possible that the change in polarity of the side chain of residue at position 196, which is at the surface of the ribose-binding region in this enzyme, facilitates the entrance of the adenosine ribose, particularly in the case of NADP+. Although many NAD+-dependent enzymes have a nonpolar residue at position 196, a strict correlation with coenzyme preference cannot be drawn, since some NADP+-dependent also have a nonpolar residue. Examples of the latter are the ALDH1L1 enzymes, but in them the 196 residues are far away from

716


701 702 703 704 705 706 707 708 709 710 711 712 713 714

717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763

CBI 7249


16 January 2015 L. González-Segura et al. / Chemico-Biological Interactions xxx (2015) xxx–xxx 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788

789 790 791 792 793 794 795 796 797 798 799 800 801 802

the ribose-binding pocket. Therefore, from the nature of the 196 residue it cannot be inferred whether it would have any effect on the coenzyme preference; knowledge of the three-dimensional structure is needed. The residue at position 249 on the helix aD of the Rossmann fold, which together with helix aC forms the cavity where the adenine ring binds, has also be substituted in order to make more room for the binding of NADP+. The Ile249Val mutant of ALDH3H1 from Arabidopsis can use NADP+, which the wild type enzyme cannot use, although the KmNADP+ of the mutant was very high. But the Val249Ile mutant of the ALDH3I1 from the same plant decreases the Km for both coenzymes compared with the values of the wild-type enzyme [20]. It seems therefore than some other factors may be involved in the observed effects of the mutations. In a following study [32], the same research group substituted Glu195 of the ALDH3H1 with Gln, Asp, Asn or Thr, increasing the affinity for NADP+, which was increased further in the double mutant Glu195Thr/Ile249Val and even further in the triple mutant Glu195Thr/Val225Arg/Ile249Val, as assessed by the decreases in KmNADP+ values and a ten-times increase in kcat/KmNADP+ when compared with the corresponding value for NAD+, although the latter value decreased about ten-times compared with the wild-type enzyme [33]. The authors concluded that an Arg at position 225, as observed in the NADP+- dependent ALDH from V. harveyi, contributes to decrease KmNADP+ by stabilizing its adenine ring through l-stacking interactions, and that the opposite effect of this mutation on KmNAD+ is due to the different orientation of the adenine ring of NAD+ provoked by the absence of the 20 -phosphate group. This explanation is consistent with what has been observed in other ALDH3 enzymes, as discussed above. Finally, the mutation of Ser195 to a Glu changed the preference from NADP+ to NAD+ in the ALDH from Synechococcus sp. PCC 7002 (unnamed ALDH family cd07100, UNIPROT code B1XMM6), and the mutation of Pro195 to Ser or Glu increased and decreased the preference for NADP+, respectively, in the fatty ALDH from Gluconobacter oxydans (unnamed ALDH family cd07100, UNIPROT code Q5FTL8) [34], as assessed by changes in the activity of these enzymes determined at fixed concentrations of substrates and coenzymes. Once again, these results agree with the crystallo-

Fig. 4. Schematic representation of the possible coenzyme preference of ALDH enzymes proposed on the basis of available structural and kinetic data. The inner circle includes the different residues at position 195; the outer semicircle includes the possible residues at position 224 in the enzymes that have Glu195. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

13

graphic evidence, and confirm the critical importance of the residue at position 195 in defining the coenzyme preference of ALDH enzymes.

803

4. Concluding remarks

806

Although some factors that may influence coenzyme preference of the ALDH enzymes have not been addressed in this work, for instance the surface electrostatic potential of the coenzyme entrance tunnel, our analyses of the published kinetic and structural data support the following general conclusions: (1) Within the ALDH superfamily, there is more diversity than it was anticipated in the structural factors relevant for coenzyme preference. (2) In spite of this, it is clear that the main determinant of coenzyme preference in many of these enzymes is the residue at position 195. (3) A Glu at position 195 generally adopts a conformation that allows its interaction with the adenosine ribose and at the same time reduces the size of the ribose-binding cavity, resulting in a preference for NAD+ over NADP+, while residues with smaller side-chains allow wider cavities and the binding of 20 -phosphate group of NADP+, but usually to the detriment of NAD+ binding. (4) Due to the particular conformation of its side chain, a Pro at position 195 favors NAD(H)-binding, as Glu does, whereas Gln195 favors NADP(H) binding, as the smaller residues do, because its side chain is not attracted by the positive charge of Lys192 amino group and therefore points in an opposite direction to that of Glu195. (5) In the NAD+-specific ALDHs that have Glu195, other residues located at a few particular positions, and not directly involved in coenzyme binding, modulate the ability of these enzymes to also bind NADP+ by permitting the movement of Glu195 side chain from its NAD+-binding position. In Fig. 4 we summarized these conclusions, showing the possible coenzyme preference of the ALDH enzymes that we propose on the basis of our analyses of their available kinetic and structural data. Surely, the kinetic characterization of more ALDH enzymes together with the determination of new crystal structures, including those of enzymes considered NAD+- or NADP+-specific in complex with the other nucleotide, will bring to light new cases and will prove or reject some of our predictions. It is interesting that in other pyridine nucleotide-linked dehydrogenases a negatively-charged residue (Asp or Glu) was suggested to perform a role similar to that of Glu195 in the NAD+specific enzymes [35], whereas the NADP+-specific ones have a neutral residue instead [36,37]. It has been however proven that replacement of this residue does not suffice to convert a NAD+-specific enzyme into a NADP+-specific [38]. In alcohol dehydrogenases the Asp residue is located at the end of the b2-strand of the Rossmann fold (see for instance the crystal structure of the horse ADH, PDB code 4DWV) while it is in the loop b2-aB in the ALDHs (Fig. 1), which reflects the differences in the Rossmann fold between these enzymes [2]. In spite of this, it is interesting that similar strategies to discriminate against NADP+ have been selected in different protein superfamilies. Our finding that coenzyme preference is not conserved within many ALDH families is evidence of the limited number of residues that influence coenzyme preference, as well as of the plasticity that is required to fulfill the physiological needs of the organism. This plasticity is mainly supported on the evolvability of residue 195 but fine-tuned by other residues or structural features, which results in a large variety of solutions to the problem of defining coenzyme preference in the ALDH superfamily. The selection of coenzyme preference could be exerted by several factors. For instance, coenzyme availability in the intracellular location of the enzyme, avoidance of competition for the coenzyme with other important reactions, or the need to provide both NADH for energy

807


804 805

808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865

CBI 7249


16 January 2015 14


878

and NADPH for biosynthetic or defense processes. The latter could be the reason for the existence of dual-coenzyme specificity bacterial ALDHs participating in the catabolism of compounds that can be used by the bacterium as the only source of carbon and energy for growth [9]. Finally, we recommend that when studying the coenzyme preference of an ALDH, in addition to determining the steady-state kinetics of the enzyme with both coenzymes, including the determination of the kinetic mechanism of the NAD+- and NADP+dependent reactions, the equilibrium binding of the coenzymes should be studied. This would allow differentiating between poor and non-productive binding, as well as the identification of the structural factors that may affect one or the other mode of binding.

879

Conflict of interest

866 867 868 869 870 871 872 873 874 875 876 877

880

[14] [15]

[16]

[17] [18]

[19]

[20]

None. [21]

881

Transparency Document

882 883 884

The Transparency document associated with this article can be found in the online version.

885

Acknowledgments

886 887 888 889 890 891

Q4 The authors acknowledge the financial support of Consejo NacQ5 ional de Ciencia y Tecnología (CONACYT Grant 167122 to R.A.M.C.) and UNAM (PAPIIT Grant IN216513 to H.R.R.). A.J.S. was supported by R13-AA023149 to present this work at the 17th Enzymology and Molecular Biology of Carbonyl Metabolism meeting held in Skytop, PA, USA.

[22]

[23]

[24]

[25]

[26] 892

References

893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932

[1] M.G. Rossmann, D. Moras, K.W. Olsen, Chemical and biological evolution of a nucleotide-binding protein, Nature 250 (1974) 194–199. [2] C.G. Steinmetz, P. Xie, H. Weiner, T.D. Hurley, Structure of mitochondrial aldehyde dehydrogenase: the genetic component of ethanol aversion, Structure 5 (1997) 701–711. [3] Z.J. Liu, Y.J. Sun, J. Rose, Y.J. Chung, C.D. Hsiao, W.R. Chang, I. Kuo, J. Perozich, R. Lindahl, J. Hempel, B.C. Wang, The first structure of an aldehyde dehydrogenase reveals novel interactions between NAD and the Rossmann fold, Nat. Struct. Biol. 4 (1997) 317–326. [4] L. Zhang, B. Ahvazi, R. Szittner, A. Vrielink, E. Meighen, Change of nucleotide specificity and enhancement of catalytic efficiency in single point mutants of Vibrio harveyi aldehyde dehydrogenase, Biochemistry 38 (1999) 11440–11447. [5] J. Perozich, I. Kuo, B.C. Wang, J.S. Boesch, R. Lindahl, J. Hempel, Shifting the NAD/NADP preference in class 3 aldehyde dehydrogenase, Eur. J. Biochem. 267 (2000) 6197–6203. [6] B. Ahvazi, R. Coulombe, M. Delarge, M. Vedadi, L. Zhang, E. Meihgen, A. Vrielink, Crystal structure of the NADP+-dependent aldehyde dehydrogenase from Vibrio harveyi: structural implications for cofactor specificity and affinity, Biochem. J. 349 (2000) 853–861. [7] D. Cobessi, F. Tete-Favier, S. Marchal, G. Branlant, A. Aubry, Structural and biochemical investigations of the catalytic mechanism of an NADP-dependent aldehyde dehydrogenase from Streptococcus mutans, J. Mol. Biol. 300 (2000) 141–152. [8] K. Johansson, M. El-Ahmad, S. Ramaswamy, L. Hjelmqvist, H. Jornvall, H. Eklund, Structure of betaine aldehyde dehydrogenase at 2.1 Å resolution, Protein Sci. 87 (1998) 2106–2117. [9] R. Velasco-García, L. González-Segura, R.A. Muñoz-Clares, Steady-state kinetic mechanism of the NADP+- and NAD+-dependent reactions catalysed by betaine aldehyde dehydrogenase from Pseudomonas aeruginosa, Biochem. J. 352 (2000) 675–683. [10] P. Emsley, K. Cowtan, Coot: model-building tools for molecular graphics, Acta Crystallogr D60 (2004) 2126–2132. [11] E. Jain, A. Bairoch, S. Duvaud, I. Phan, N. Redaschi, B.E. Suzek, M.J. Martin, P. McGarv, E. Gasteiger, Infrastructure for the life sciences: design and implementation of the UniProt website, BMC Bioinformatics 10 (2009) 136. [12] K. Tamura, G. Stecher, D. Peterson, A. Filipski, S. Kumar, MEGA6: Molecular evolutionary genetics analysis version 6.0, Mol. Biol. Evol. 30 (2013) 2725– 2729. [13] M.A. Larkin, G. Blackshields, N.P. Brown, R. Chenna, P.A. McGettigan, H. McWilliam, F. Valentin, I.M. Wallace, A. Wilm, R. Lopez, J.D. Thompson, T.J.

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36] [37]

[38]

Gibson, D.G. Higgins, Clustal W and clustal X version 2.0, Bioinformatics 23 (2007) 2947–2948. G.E. Crooks, G. Hon, J.M. Chandonia, S.E. Brenner, WebLogo: a sequence logo generator, Genome Res. 14 (2004) 1188–1190. S.A. Krupenko, C. Wagner, R.J. Cook, Expression, purification, and properties of the aldehyde dehydrogenase homologous carboxyl-terminal domain of rat 10formyltetrahydrofolate dehydrogenase, J. Biol. Chem. 272 (1997) 10266– 10272. R.I. Feldman, H. Weiner, Horse liver aldehyde dehydrogenase. II. Kinetics and mechanistic implications of the dehydrogenase and esterase activity, J. Biol. Chem. 247 (1972) 267–272. A.J. Rivett, K.F. Tipton, Kinetic studies with rat-brain succinic-semialdehyde dehydrogenase, Eur. J. Biochem. 117 (1981) 187–193. E.M. Valenzuela-Soto, R.A. Muñoz-Clares, Betaine-aldehyde dehydrogenase from leaves of Amaranthus hypochondriacus L. exhibits an Iso Ordered Bi Bi steady state mechanism, J. Biol. Chem. 268 (1993) 23818–23823. R.A. Muñoz-Clares, C. Mújica-Jiménez, Complexes of NADH with betaine aldehyde dehydrogenase from leaves of the plant Amaranthus hypochondriacus L, Chem.-Biol. Interact. 120–132 (2001) 71–80. N. Stiti, I.O. Adewale, J. Petersen, D. Bartels, H.-H. Kirch, Engineering the nucleotide coenzyme specificity and sulfhydryl redox sensitivity of two stressresponsive aldehyde dehydrogenase isoenzymes of Arabidopsis thaliana, Biochem. J. 434 (2011) 459–471. W. Black, V. Vasiliou, The aldehyde dehydrogenase gene superfamily resource center, Hum. Genomics 4 (2009) 136–142. X. Peng, K. Shindo, K. Kanoh, Y. Inomata, S.-K. Choi, N. Misawa, Characterization of Sphingomonas aldehyde dehydrogenase catalyzing the conversion of various aromatic aldehydes to their carboxylic acids, Appl. Microbiol. Biotechnol. 69 (2005) 141–150. H. Riveros-Rosas, L. González-Segura, A. Julián-Sánchez, A.G. Díaz-Sánchez, R.A. Muñoz-Clares, Structural determinants of substrate specificity in aldehyde dehydrogenases, Chem.-Biol. Interact. 202 (2013) 51–61. A. Marchler-Bauer, S. Lu, J.B. Anderson, F. Chitsaz, M.K. Derbyshire, C. DeWeese-Scott, J.H. Fong, L.Y. Geer, R.C. Geer, N.R. Gonzales, M. Gwadz, D.I. Hurwitz, J.D. Jackson, Z. Ke, C.J. Lanczycki, F. Lu, G.H. Marchler, M. Mullokandov, M.V. Omelchenko, C.L. Robertson, J.S. Song, N. Thanki, R.A. Yamashita, D. Zhang, N. Zhang, C. Zheng, S.H. Bryant, CDD: a conserved domain database for the functional annotation of proteins, Nucleic Acids Res. 39 (2011) D225–D229. C. Chen, J.C. Joo, G. Brown, E. Stolnikova, A.S. Halavaty, A. Savchenko, W.F. Anderson, A.F. Yakunin, Substrate inhibition of the betaine aldehyde dehydrogenase BetB from Staphylococcus aureus: structure-based mutational studies, Appl. Environ. Microbiol. 80 (2014) 3992–4002. L. González-Segura, E. Rudiño-Piñera, R.A. Muñoz-Clares, E. Horjales, The crystal structure of a ternary complex of betaine aldehyde dehydrogenase from Pseudomonas aeruginosa provides a new insight into the reaction mechanism and shows a novel binding mode of the 20 -phosphate of NADP+ and a novel cation binding site, J. Mol. Biol. 385 (2009) 542–557. M.J. McLeish, M.M. Kneen, K.N. Gopalakrishna, C.W. Koo, P.C. Babbitt, J.A. Gerlt, G.L. Kenyon, Identification and characterization of a mandelamide hydrolase and an NAD(P)+-dependent benzaldehyde dehydrogenase from Pseudomonas putida ATCC 12633, J. Bacteriol. 185 (2003) 2451–2456. E. Inagaki, N. Ohshima, H. Takahashi, C. Kuroishi, S. Yokoyama, T.H. Tahirov, Crystal structure of Thermus thermophilus D1-pyrroline-5-carboxylate dehydrogenase, J. Mol. Biol. 362 (2006) 490–501. M. Vedadi, A. Vrielink, E. Meighen, Involvement of conserved glycine residues, 229 and 234, of Vibrio harveyi aldehyde dehydrogenase in activity and nucleotide binding, Biochem. Biophys. Res. Commun. 238 (1997) 448–451. N.A. Brunner, H. Brinkmann, B. Siebers, R. Hensel, NAD+-dependent glyceraldehyde-3-phosphate dehydrogenase from Thermoproteus tenax: The first identified archaeal member of the aldehyde dehydrogenase superfamily is a glycolytic enzyme with unusual regulatory properties, J. Biol. Chem. 273 (1998) 6149–6156. E. Lorentzen, R. Hensel, T. Knura, H. Ahmed, E. Pohl, Structural basis of allosteric regulation and substrate specificity of the non-phosphorylating glyceraldehyde 3-phosphate dehydrogenase from Thermoproteus tenax, J. Mol. Biol. 341 (2004) 815–828. J.S. Rodríguez-Zavala, Enhancement of coenzyme binding by a single point mutation at the coenzyme binding domain of E. coli lactaldehyde dehydrogenase, Protein Sci. 17 (2008) 563–570. N. Stiti, K. Podgórska, D. Bartels, Aldehyde dehydrogenase enzyme ALDH3H1 from Arabidopsis thaliana: identification of amino acid residues critical for cofactor specificity, Biochem. Biophys. Acta 2014 (1844) 681–693. Z. Yuan, B. Yin, D. Wei, Y.A. Yuan, Structural basis for cofactor and substrate selection by cyanobacterium succinic semialdehyde dehydrogenase, J. Struct. Biol. 182 (2013) 125–135. I. Hanukoglu, T. Gutfinger, CDNA sequence of adrenodoxin reductase. Identification of NADP-binding sites in oxidoreductases, Eur. J. Biochem. 180 (1989) 479–484. N.S. Scrutton, A. Berry, R.N. Perham, Redesign of the coenzyme specificity of a dehydrogenase by protein engineering, Nature 343 (1990) 38–43. P.J. Baker, K.L. Britton, D.W. Rice, A. Rob, T.J. Stillman, Structural consequences of sequence patterns in the fingerprint region of the nucleotide binding fold. Implications for nucleotide specificity, J. Mol. Biol. 228 (1992) 662–671. F. Fan, B.V. Plapp, Probing the affinity and specificity of yeast alcohol dehydrogenase I for coenzymes, Arch. Biochem. Biophys. 367 (1999) 240–249.


933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018

CBI 7249


16 January 2015 L. González-Segura et al. / Chemico-Biological Interactions xxx (2015) xxx–xxx 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090

1091 1092 1093 1094 1095 1096 1097 1098

[39] K.K. Ho, A. Allali-Hassani, T.D. Hurley, H. Weiner, Differential effects of Mg2+ ions on the individual kinetic steps of human cytosolic and mitochondrial aldehyde dehydrogenases, Biochemistry 44 (2005) 8022–8029. [40] D. Stagos, Y. Chen, C. Brocker, E. Donald, B.C. Jackson, D.J. Orlicky, D.C. Thompson, V. Vasiliou, Aldehyde dehydrogenase 1B1: Molecular cloning and characterization of a novel mitochondrial acetaldehyde-metabolizing enzyme, Drug Metab. Dispos. 38 (2010) 1679–1687. [41] A.K.H. MacGibbon, L.F. Blackwell, P.D. Buckley, Kinetics of sheep-liver cytoplasmic aldehyde dehydrogenase, Eur. J. Biochem. 77 (1977) 93–100. [42] X. Wang, P. Penzes, J.L. Napoli, Cloning of a cDNA encoding an aldehyde dehydrogenase and its expression in Escherichia coli. Recognition of retinal as substrate, J. Biol. Chem. 271 (1996) 16288–16293. [43] C.E. Graham, K. Brocklehurst, R.W. Pickersgill, M.J. Warren, Characterization of retinaldehyde dehydrogenase 3, Biochem. J. 394 (2006) 67–75. [44] X. Wang, H. Weiner, Involvement of glutamate 268 in the active site of human liver mitochondrial (class 2) aldehyde dehydrogenase as probed by sitedirected mutagenesis, Biochemistry 34 (1995) 237–243. [45] W. Leicht, F. Heinz, B. Freimüller, Purification and characterization of aldehyde dehydrogenase from bovine liver, Eur. J. Biochem. 83 (1978) 189–196. [46] R. Lindhal, S. Evces, Rat liver aldehyde dehydrogenase. I. Isolation and characterization of four high Km normal liver isoenzymes, J. Biol. Chem. 259 (1984) 11986–11990. [47] B. Rothacker, T. Ilg, Functional characterization of a Drosophila melanogaster succinic semialdehyde dehydrogenase and a non-specific aldehyde dehydrogenase, Insect Biochem. Mol. 38 (2008) 354–366. [48] X. Wang, C.J. Mann, Y. Bai, L. Ni, H. Weiner, Molecular cloning, characterization, and potential roles of cytosolic and mitochondrial aldehyde dehydrogenases in ethanol metabolism in Saccharomyces cerevisiae, J. Bacteriol. 180 (1998) 822– 830. [49] L.C. Hsu, W.-C. Chang, A. Shibuya, A. Yoshida, Human stomach aldehyde dehydrogenase cDNA and genomic cloning, primary structure, and expression in Escherichia coli, J. Biol. Chem. 267 (1992) 3030–3037. [50] H.-H. Kirch, A. Nair, D. Bartels, Novel ABA- and dehydration-inducible aldehyde dehydrogenase genes isolated from the resurrection plant Craterostigma plantagineum and Arabidopsis thaliana, Plant J. 28 (2001) 555– 567. [51] D. Srivastava, R.K. Singh, M.A. Moxley, M.T. Henzl, D.F. Becker, J.J. Tanner, The three-dimensional structural basis of type II hyperprolinemia, J. Mol. Biol. 420 (2012) 176–189. [52] B.W. Arentson, M. Luo, T.A. Pemberton, J.J. Tanner, D.F. Becker, Kinetic and structural characterization of tunnel-perturbing mutants in Bradyrhizobium japonicum proline utilization A, Biochemistry 53 (2014) 5150–5161. [53] W.S. Blaner, J.E. Churchich, The binding of NADH to succinic semialdehyde dehydrogenase, Eur. J. Biochem. 109 (1980) 431–437. [54] K.B. Busch, H. Fromm, Plant succinic semialdehyde dehydrogenase. Cloning, purification, localization in mitochondria, and regulation by adenine nucleotides, Plant Physiol. 121 (1999) 589–597. [55] I. Yamaura, T. Matsumoto, M. Funatsu, T. Shinohara, Purification and some properties of succinic semialdehyde dehydrogenase from barley seeds, Agric. Biol. Chem. 52 (1988) 2929–2930. [56] V.S. Narayan, P.M. Nair, Potato tuber succinate semialdehyde dehydrogenase: Purification and characterization, Arch. Biochem. Biophys. 275 (1989) 469– 477. [57] M. Jaeger, B. Rothacker, T. Ilg, Saturation transfer difference NMR studies on substrates and inhibitors of succinic semialdehyde dehydrogenases, Biochem. Biophys. Res. Commun. 372 (2008) 400–406. [58] S.A. Park, Y.S. Park, K.S. Lee, Kinetic characterization and molecular modeling of NAD(P)+-dependent succinic semialdehyde dehydrogenase from Bacillus subtilis as an ortholog YneI, J. Microbiol. Biotechnol. 24 (2014) 954–958. [59] G.W. Goodwin, P.M. Rougraff, E.J. Davis, R.A. Harris, Purification and characterization of methylmalonate-semialdehyde dehydrogenase from rat liver, J. Biol. Chem. 264 (1989) 14965–14971. [60] C.C. Shone, H.J. Fromm, Steady-state and pre-steady-state kinetics of coenzyme A linked aldehyde dehydrogenase from Escherichia coli, Biochemistry 20 (1981) 7494–7501. [61] D. Bannerjee, L.E. Sanders, J.R. Sokatch, Properties of purified methylmalonate semialdehyde dehydrogenase of Pseudomonas aeruginosa, J. Biol. Chem. 245 (1970) 1828–1835. [62] C. Stines-Chaumeil, F. Talfournier, G. Branlant, Mechanistic characterization of the MSDH (methylmalonate semialdehyde dehydrogenase) from Bacillus subtilis, Biochem. J. 395 (2006) 107–115. [63] Y.-F. Chang, P. Ghosh, V.V. Rao, L-pipecolic acid metabolism in human liver: La aminoadipate d-semialdehyde oxidoreductase, Biochim. Biophys. Acta 1038 (1990) 300–305. [64] W.-K. Tang, C.H.K. Cheng, W.-P. Fong, First purification of the antiquitin protein and demonstration of its enzymatic activity, FEBS Lett. 516 (2002) 183–186. [65] C.-H. Tsai, L.M. Henderson, Degradation of O-phosphohydroxylysine in rat liver: purification and properties of 2-aminoadipic semialdehyde dehydrogenase, J. Biol. Chem. 249 (1974) 5790–5792.

15

[66] G. Kurys, W. Ambroziak, R. Pietruszko, Human aldehyde dehydrogenase: purification and characterization of a third isozyme with low Km for caminobutyraldehyde, J. Biol. Chem. 264 (1989) 4715–4721. [67] A.M. Guzman-Partida, E.M. Valenzuela-Soto, Porcine kidney betaine aldehyde dehydrogenase: purification and properties, Comp. Biochem. Physiol. 119 (1998) 485–491. [68] F.M. Vaz, S.W. Fouchier, R. Ofman, M. Sommer, R.J.A. Wanders, Molecular and biochemical characterization of rat c-trimethylaminobutyraldehyde dehydrogenase and evidence for the involvement of human aldehyde dehydrogenase 9 in carnitine biosynthesis, J. Biol. Chem. 275 (2000) 7390– 7394. [69] P. Falkenberg, A.R. Strom, Purification and characterization of osmoregulatory betaine aldehyde dehydrogenase of Escherichia coli, Biochim. Biophys. Acta 1034 (1990) 253–259. [70] N. Mori, S. Fuchigami, Y. Kitamoto, Purification and properties of betaine aldehyde dehydrogenase from Xanthomonas translucens, J. Ferment. Bioeng. 73 (1992) 350–356. [71] E.A. Weretilnyk, A.D. Hanson, Molecular cloning of a plant betaine-aldehyde dehydrogenase, an enzyme implicated in adaptation to salinity and drought, Proc. Natl. Acad. Sci. USA 87 (1990) 2745–2749. [72] M. Tylichová, D. Kopecˇny´, S. Moréra, P. Briozzo, R. Lenobel, J. Snégaroff, M. Šebela, Structural and functional characterization of plant aminoaldehyde dehydrogenase from Pisum sativum with a broad specificity for natural and synthetic aminoaldehydes, J. Mol. Biol. 396 (2010) 870–882. [73] D. Kopecˇny´, R. Koncˇitíková, M. Tylichová, A. Vigouroux, H. Moskalíková, M. Soural, M. Šebela, S. Moréra, Plant ALDH10 Family: Identifying critical residues for substrate specificity and trapping a thiohemiacetal intermediate, J. Biol. Chem. 288 (2013) 9491–9507. [74] J.R. Livingstone, I. Yoshida, Y. Tarui, K. Hirooka, Y. Yamamoto, N. Tsutui, E. Hirasawa, Purification and properties of aminoaldehyde dehydrogenase from Avena sativa, J. Plant. Res. 115 (2002) 393–400. [75] M. Hassan, M. Okada, T. Ichiyanagi, N. Mori, 4-Ntrimethylaminobutyraldehyde dehydrogenase: purification and characterization of an enzyme from Pseudomonas sp. 13CM, Biosci. Biotechnol. Biochem. 72 (2008) 155–162. [76] S. Marchal, G. Branlant, Evidence for the chemical activation of essential Cys302 upon cofactor binding to nonphosphorylating glyceraldehyde 3phosphate dehydrogenase from Streptococcus mutans, Biochemistry 38 (1999) 12950–12958. [77] T.J.G. Ettema, H. Ahmed, A.C.M. Geerling, J. van der Oost, B. Siebers, The nonphosphorylating glyceraldehyde-3-phosphate dehydrogenase (GAPN) of Sulfolobus solfataricus: a key-enzyme of the semi-phosphorylative branch of the Entner–Doudoroff pathway, Extremophiles 12 (2008) 75–88. [78] K.K. Ho, H. Weiner, Isolation and characterization of an aldehyde dehydrogenase encoded by the aldB gene of Escherichia coli, J. Bacteriol. 187 (2005) 1067–1073. [79] A. Jaureguibeitia, L. Saá, M.J. Llama, J.L. Serra, Purification, characterization and cloning of aldehyde dehydrogenase from Rhodococcus erythropolis UPV-1, Appl. Microbiol. Biotechnol. 73 (2007) 1073–1086. [80] J. Boch, G. Nau-Wagner, S. Kneip, E. Bremer, Glycine betaine aldehyde dehydrogenase from Bacillus subtilis: characterization of an enzyme required for the synthesis of the osmoprotectant glycine betaine, Arch. Microbiol. 168 (1997) 282–289. [81] N.N. Samsonova, S.V. Smirnov, A.E. Novikova, L.R. Ptitsyn, Identification of Escherichia coli K12 YdcW protein as a c-aminobutyraldehyde dehydrogenase, FEBS Lett. 579 (2005) 4107–4112. [82] J. Bains, M.J. Boulanger, Structural and biochemical characterization of a novel aldehyde dehydrogenase encoded by the benzoate oxidation pathway in Burkholderia xenovorans LB400, J. Mol. Biol. 379 (2008) 597–608. [83] M. Vedadi, R. Szittner, L. Smillie, E. Meighen, Involvement of cysteine 289 in the catalytic activity of an NADP+-specific fatty aldehyde dehydrogenase from Vibrio harveyi, Biochemistry 34 (1995) 16725–16732. [84] J. Park, S. Rhee, Structural basis for a cofactor-dependent oxidation protection and catalysis of cyanobacterial succinic semialdehyde dehydrogenase, J. Biol. Chem. 288 (2013) 15760–15770. [85] H. Zheng, A. Beliavsky, A. Tchigvintsev, J.S. Brunzelle, G. Brown, R. Flick, E. Evdokimova, Z. Wawrzak, R. Mahadevan, W.F. Anderson, A. Savchenko, A.F. Yakunin, Structure and activity of the NAD(P)+-dependent succinate semialdehyde dehydrogenase YneI from Salmonella typhimurium, Proteins 81 (2013) 1031–1041. [86] P. Schweiger, S. Volland, U. Deppenmeier, Overproduction and characterization of two distinct aldehyde-oxidizing enzymes from Gluconobacter oxydans 621H, J. Mol. Microbiol. Biotechnol. 13 (2007) 147–155. [87] T. Liu, L. Hao, R. Wang, B. Liu, Molecular characterization of a thermostable aldehyde dehydrogenase (ALDH) from the hyperthermophilic archaeon Sulfolobus tokodaii strain 7, Extremophiles 17 (2013) 181–190. [88] J.S. Rodríguez-Zavala, A. Allali-Hassani, H. Weiner, Characterization of E. coli tetrameric aldehyde dehydrogenases with atypical properties compared to other aldehyde dehydrogenases, Protein Sci. 15 (2006) 1387–1396.


1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178

Aldehyde dehydrogenases and their role in carcinogenesis.

Inducible aldehyde dehydrogenases in the hepatic cytosol of the rat.

Identification of mouse liver aldehyde dehydrogenases that catalyze the oxidation of retinaldehyde to retinoic acid.

Identification of human liver aldehyde dehydrogenases that catalyze the oxidation of aldophosphamide and retinaldehyde.

Aldehyde dehydrogenases in cancer stem cells: potential as therapeutic targets.

Aldehyde dehydrogenases in early stage lung cancer: nuclear expression.

Localization and characteristics of rat liver mitochondrial aldehyde dehydrogenases.

Purification and characterization of two aldehyde dehydrogenases from Pseudomonas aeruginosa.

Lipid aldehyde oxidation as a physiological role for class 3 aldehyde dehydrogenases.

A novel coenzyme from bacterial primary alcohol dehydrogenases.

KAHA ligations that form aspartyl aldehyde residues as synthetic handles for protein modification and purification.

The role of ala198 in the stability and coenzyme specificity of bacterial formate dehydrogenases.

Comparison of phenobarbital- and carcinogen-induced aldehyde dehydrogenases in the rat.

The effect of some analogues of disulfiram on the aldehyde dehydrogenases of sheep liver.

alcohol dehydrogenases to butanol and ethanol production in Clostridium acetobutylicum.

The effect of disulfiram on the aldehyde dehydrogenases of sheep liver.

The effect of ethanol ingestion on the aldehyde dehydrogenases of rat liver.

Two aldehyde dehydrogenases from human liver. Isolation via affinity chromatography and characterization of the isozymes.

Functional arginine residues involved in coenzyme binding by dihydrofolate reductase.

Meta-analysis of the aldehyde dehydrogenases-2 (ALDH2) Glu487Lys polymorphism and colorectal cancer risk.

Partial reversal of the acetaldehyde and butyraldehyde oxidation reactions catalysed by aldehyde dehydrogenases from sheep liver.

Alcohol and Aldehyde Dehydrogenases Contribute to Sex-Related Differences in Clearance of Zolpidem in Rats.

Identification of reactive tyrosine residues in cysteine-reactive dehydrogenases. Differences between liver sorbitol, liver alcohol and Drosophila alcohol dehydrogenases.

Identification and characterisation of Aedes aegypti aldehyde dehydrogenases involved in pyrethroid metabolism.