- Email: [email protected]
Aldo–keto reductases as modulators of stress response
Chemico-Biological Interactions 143 /144 (2003) 325 /332 www.elsevier.com/locate/chembioint
Aldo keto reductases as modulators of stress response /
Qing Chang, Theresa M. Harter, Loryn T. Rikimaru, J. Mark Petrash * Departments of Ophthalmology and Visual Sciences and of Genetics, Washington University School of Medicine, 660 S. Euclid Avenue, Campus Box 8096, St. Louis, MO 63110, USA
Abstract Human aldose reductase (AKR1B1) has been implicated as a factor in the pathogenesis of diabetic complications. However, little is known about the physiological role of this enzyme or of related aldo /keto reductases in human tissues. In mammalian systems, a gene knock out approach is often employed as an experimental strategy to probe for gene function. However, in the murine system, phenotypic characterization of an aldose reductase (AKR1B3) knock out is likely to be complicated due to functional compensation by redundant AKRs including AKRs 1A (aldehyde reductase), 1B7 (FR-1) and 1B8 (MVDP). As an alternate strategy, we are examining the budding yeast Saccharomyces cerevisiae as a model system for a functional genomics study of AKRs. A distinct advantage of this system centers on the ability to readily ablate multiple targeted genes in a single strain. In addition to providing insights into functional redundancy, this system allows us to use a genetic approach to study possible effector pathways associated with one or more individual genes. Yeast open reading frames (ORFs) encoding AKRs with functional similarity to human aldose reductase (AKR1B1) were identified by BLAST analysis and were functionally validated by studies of recombinant proteins. By ablating three of the yeast AKR genes most functionally similar to AKR1B1, we have created a unique strain of S. cerevisiae that shows enhanced sensitivity to stress. Ongoing studies with oligonucleotide arrays show that the triple null strain has an altered transcription profile consistent with an enhanced stress response in comparison with the parental strain. These data indicate that AKR-null strains may provide new insights into signaling mechanisms involving this family of proteins. # 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Aldo /keto reductase; Saccharomyces cerevisiae ; Gene chip; Stress response; Suppressor
1. Introduction Diabetes mellitus is recognized as a leading cause of new cases of blindness among Americans between the ages of 20 and 74 years. In comparison with the general population, diabetic patients
* Corresponding author. Tel.: /1 (314) 362-3335; fax: /1 (314) 362-3638 E-mail address: [email protected] (J.M. Petrash).
face a 25-fold increased risk of blindness resulting from diabetic retinopathy or cataract. While many theories have been advanced to explain mechanisms leading to diabetic complications, a prominent model suggests that enhanced flux of glucose through the polyol pathway may be causally linked to diabetic complications. Two enzymes make up the polyol pathway: aldose reductase (AKR1B) and sorbitol dehydrogenase, a short chain dehydrogenase/reductase. Evidence linking enhanced polyol pathway activity and diabetic
0009-2797/02/$ - see front matter # 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 0 0 9 - 2 7 9 7 ( 0 2 ) 0 0 2 1 6 - 8
326
Q. Chang et al. / Chemico-Biological Interactions 143 /144 (2003) 325 /332
complications comes mostly from animal studies where severity of disease can be altered by pharmacologic blockade [1] or genetic enhancement [2] of aldose reductase activity. Despite success in preventing complications in experimentally diabetic animal models treated with aldose reductase inhibitors, translation of this strategy to human therapy has not been successful [3]. Various explanations have been advanced to rationalize this discordance [4]. Perhaps of prime importance, the Diabetes Control and Complications Trial provided compelling evidence that early intervention is an important element in successfully arresting or preventing the onset and/or progression of diabetic complications [5]. Despite the substantial investment of human and economic capital in the study of aldose reductase inhibition, very little is known about the physiological role of this enzyme. The most abundant tissue source of aldose reductase is the adrenal gland, but its physiological role in that specialized tissue is not known [6]. Human tissues contain multiple aldo /keto reductases with structural and catalytic properties strikingly similar to aldose reductase. Newly discovered aldose reductase-like proteins include the human small intestine reductase AKR1B10 [7] (also known as human aldose reductase-like protein 1 [8]). Kinetic studies of aldose reductase and related AKRs point to a possible role for these enzymes in aldehyde detoxification [9 /11]. However, functional proof that aldose reductase fulfills this role in vivo has been limited to the ischemia /reperfusion process in the cardiovascular system [12]. Ablation of a targeted gene can often provide useful insight into possible physiological roles, but functional redundancy among similar aldo /keto reductases complicates this strategy in available animal models. Indeed, knock out mice resulting from ablation of the mouse aldose reductase gene were without a meaningful phenotype that could be related to aldose reductase function outside the kidney [13]. As an alternative to gene ablation in the murine animal model, we began an investigation into use of Saccharomyces cerevisiae as a model system.
1.1. Multiplicity of yeast aldo /keto reductases Six yeast open reading frames (ORFs) that encode polypeptides with sequence similarity to human aldose reductase (AKR1B1) can be identified by BLAST analysis. In Fig. 1, the encoded polypeptides are shown aligned with the AKR1B1 sequence. From the sequence alignment, it can be seen that all of the putative yeast AKRs are of a molecular mass very similar to the human enzyme. Further, it can be appreciated that all have conserved those amino acid side chains that were previously demonstrated to be of key importance in the catalytic activity of mammalian AKRs, including aldose reductase (AKR1B1), aldehyde reductase (AKR1A), and several of the hydroxysteroid dehydrogenases (e.g. AKR 1C1) [14]. These side chains include Asp-43, Tyr-48, Lys-88 and His-110 (numbered according to the human aldose reductase primary sequence). Previous studies have shown that some but not all of the proteins encoded by the putative yeast AKR ORFs can function as active enzymes. Table 1 shows the AKR ORFs, their conventional AKR designations, and any associated functional nomenclature from the published literature. Proteins derived from the six yeast AKRs can be sorted into three subfamilies of the AKR protein superfamily (see http://www.med.upenn.edu/akr/): YHR104Wp (AKR2B6) is categorized into subfamily 2 along with xylose reductases from other yeast species. Polypeptides encoded by three ORFs can be sorted into AKR subfamily 3. These include YOR120Wp (AKR3A1), YDR368Wp (AKR3A2) and YBR149Wp (AKR3C). Subfamily 5 contains two of the putative yeast AKRs, including YJR096Wp (AKR5F) and YDL124Wp (AKR5G). Interestingly, proteins from subfamilies 2 and 3 (YHR104W, YDR368W, and YBR149W) were initially identified on the basis of their enzymatic functions [15 /17]. Thus, YBR149W encodes one subunit of the heterodimeric enzyme arabinose dehydrogenase gene (ARA1 ) [17], while YDR368W is synonymous with the YPR1 gene, which encodes a xylose reductase [15,16]. The gene encoding YHR104Wp was discovered on the basis of its induction in response to stress and was named GRE3 (genes de respuestra a estres) [18];
Q. Chang et al. / Chemico-Biological Interactions 143 /144 (2003) 325 /332
327
Fig. 1. Predicted sequences of yeast aldo /keto reductases. Yeast sequences are identified by their seven character ORF names. Alignments with the primary sequence of human aldose reductase (HAR) were edited using GeneDoc (http://www.psc.edu/biomed/ genedoc/). Amino acids shown in previous studies to play catalytic roles in human aldo /keto reductases are identified by shading.
additional studies demonstrated aldo /keto reductase activity associated with YHR104Wp [15]. While initial discovery of GCY1 (galactose-inducible crystallin-like protein) centered on its induc-
tion by galactose [19], we and others subsequently demonstrated that Gcy1p is a relatively efficient aldo /keto reductase [20,21]. To date, no studies have implicated a functional role for YJR096Wp,
328
Q. Chang et al. / Chemico-Biological Interactions 143 /144 (2003) 325 /332
Table 1 Aldo /keto reductases in S. cerevisiae ORF designation AKR designation Gene name Reported function
YHR104W AKR2B6 GRE3 Aldo /keto reductase
YOR120W AKR3A1 GCY1 Aldo /keto reductase
YDR368W AKR3A2 YPR1 Xylose reductase
YBR149W AKR3C ARA1 Arabinose dehydrogenase
YJR096W AKR5F
YDL124W AKR5G
although we previously demonstrated this protein to have aldo /keto reductase activity with a variety of substrates [20]. As we reported previously, the yeast system provides many advantages for functional characterization of genes comprising a complex gene family [20]. Since our studies ultimately seek to establish a physiological role for human aldose reductase, we hypothesized that the sequence of yeast AKRs selected for evaluation should be determined by the degree of structural and catalytic similarity to the human enzyme. Based in part on kinetic and structural considerations summarized in our previously published work [20], we have chosen to evaluate the yeast AKR genes in the following order of priority: YPR1, GRE3, GCY1, YJR096W, YDL124W, ARA1. Work presented in the present paper focuses exclusively on YPR1, GRE3 and GCY1.
deletion strains would be phenotypically indistinguishable from wild type due to functional compensation by remaining AKR genes. Indeed, in phenotypic screens for temperature dependence on growth, carbon auxotrophy and morphology, the deletion strains comprising either one or two AKR gene disruptions were indistinguishable from the parental strain. In contrast, the triple AKR null strain ypr1 Dgre3 Dgcy1 D showed a marked difference from the wild type strain in heat shock sensitivity (Fig. 2). It is interesting to note that the threshold for emergence of a clear phenotype appears to lie at three AKR genes, as deletion of two AKR genes, in any combination, fails to give rise to a measurable phenotype, whereas deletion of three AKR genes gives rise to a heat shock phenotype. The presence of any one of the three catalytically similar yeast AKRs appears to be sufficient to confer protection against heat shock sensitivity.
2. Results and discussion
2.2. Gene chip studies
2.1. Gene disruption studies
As an additional strategy to characterize the triple null AKR mutant, we carried out transcription profiling studies with the use of oligonucleotide arrays. To ascertain changes in gene expression under basal and heat shock conditions, total RNA was extracted from triplicate wild type and triple AKR null cultures. For heat shock treatment, mid-log phase cultures at 30 8C were shifted to 37 8C for 30 min. Cultures under basal conditions were maintained at 30 8C. Oligonucleotide array chips were run in triplicate for each condition and for each yeast strain, resulting in a total of 12 gene chips in this analysis. Array data were analyzed using the AFFYMETRIX MICROARRAY SUITE (v5.0) of programs and GENESPRING (Silicon Genetics) for filtering and pathway mapping. We established a cutoff of 2.5-fold as the
A series of isogenic yeast strains was constructed to represent all possible permutations of yeast AKR gene deletions involving YPR1, GRE3 and GCY1. Thus, strains were constructed with a single AKR gene deletion (ypr1 D, gre3 D, gcy1 D), as well as strains with two AKR genes deleted (ypr1 D gre3 D; ypr1 Dgcy1 D; gre3 Dgcy1 D). Finally, a triple AKR null strain was developed (ypr1 Dgre3 Dgcy1 D). Each strain was analyzed for phenotypic changes, utilizing a battery of tests selected to screen for changes in temperature sensitivity, carbon auxotrophy, morphology and stress response. Given the broad overlap in functional similarity, at least at the level of aldehyde reduction, we predicted that single AKR gene
Q. Chang et al. / Chemico-Biological Interactions 143 /144 (2003) 325 /332
329
Fig. 2. Heat shock sensitivity of triple AKR null strain. Serially-diluted cultures grown under control conditions or following heat shock treatment (55 8C, 2 min) were spotted on agar plates for growth at 30 8C. In comparison with the wild type strain, cultures of the triple AKR null strain are substantially diminished following heat shock treatment. Fig. 3. Venn diagram of genes responding to heat shock. Wild type cells responded with 448 genes showing ]/2.5-fold increase in transcripts (red and purple) while the triple AKR null strain responded with 61 genes showing a ]/2.5-fold increase in transcripts (blue and purple). A comparison of the genes identified from the two strains showed that 50 were in common (purple). Fig. 4. Venn diagram of wild type and triple null strains. Wild type cells responded with 448 genes showing ]/2.5-fold increase in transcripts (red and yellow). Comparison of the triple AKR null strain with wild type grown under basal conditions shows that the triple null strain has 241 genes that are constitutively upregulated by a factor of ]/2.5-fold (green and yellow). A comparison of the genes identified from the two growth conditions shows that 143 were in common (yellow).
threshold to define what we consider to be a meaningful change in transcript level. Examination of wild type cells revealed that heat shock resulted in the induction of 448 genes (Fig. 3). Similar examination of the triple AKR null cells showed that only 61 genes were induced by heat shock treatment. A comparison of the induced genes identified in the WT and triple AKR
null strain showed that most of the heat shock induced genes in the triple null strain were also induced by heat shock in the WT strain (Fig. 3). This demonstrates that the capability to mount a heat shock response is largely preserved in the triple AKR null strain. A comparison of WT and triple null strains cultured under basal conditions revealed that 241
330
Q. Chang et al. / Chemico-Biological Interactions 143 /144 (2003) 325 /332
Fig. 5. Venn diagram of heat shock response in wild type and triple null strains. Area of circle marked by RED represents induced genes in wild type cells subjected to heat shock. Area of circle marked by GREEN represents induced genes in triple AKR null cells grown under basal conditions. Area of circle marked by BLUE represents induced genes in triple AKR null cells subjected to heat shock. The area representing overlap of all three circles shows that 15 genes in common were upregulated in all comparison sets. Fig. 6. Functional categories represented by constitutively upregulated genes in the triple AKR null strain.
Q. Chang et al. / Chemico-Biological Interactions 143 /144 (2003) 325 /332
331
Table 2 Specificity of transcriptional remodeling Category
# Genes induced
Total # ORFS in category
Percentage induced
Energy/metabolism Gene expression Cellular organization Transport Signaling/interactions Cell cycle/DNA processing Unclassified Total
36 34 47 14 7 30 43 215
1318 1725 2258 808 467 906 2399 10 556
2.7 2.0 2.1 1.7 1.5 3.3 1.8 2.0
genes were induced at least 2.5-fold in the triple null strain. As shown in the Venn diagram in Fig. 4, approximately 60% of the constitutively-induced genes in the triple AKR null strain (143 of 241) were identified as heat shock responsive genes in the WT. Indeed, of the 143 constitutively upregulated genes in the triple null strain that were heat shock induced in the WT, only 15 (10%) were further upregulated by heat shock in the triple AKR null strain (Fig. 5). With the use of pathway mapping tools, we sorted each of the constitutively upregulated genes into functional categories. Fig. 6 shows that upregulated genes could be sorted into major categories involved in metabolism, gene expression, cellular organization and transport. Table 2 shows that constitutively induced genes in the triple AKR null strain comprise approximately 2% of all genes in the sample pool. Thus, it appears that the transcriptome of the AKR triple null strain reflects constitutive activation of many stress response genes. This suggests a potential role for aldo /keto reductases in downregulation of genes that are controlled by the stress response pathway. Since many stress response genes are activated through the action of Msn2/ Msn4 transcriptional factors [22], we hypothesize that the aldo /keto reductases identified in the present study could act as signaling modulators along the pathway leading to regulation of Msn2p and/or Msn4p. In this context, lack of sufficient AKR functionality could result in constitutive activation of Msn2p/Msn4p, leading to dysregulated transcription of stress response genes.
Acknowledgements This work was supported in part from NIH grants EY05856, EY02687, and 2P60DK20579. Additional support was provided by Research to Prevent Blindness, Inc. through a Lew R. Wasserman Award (J.M. Petrash) and an unrestricted award to the Washington University Department of Ophthalmology and Visual Sciences. The support of the Kilo Diabetes and Vascular Research Foundation is also gratefully acknowledged.
References [1] J.H. Kinoshita, C. Nishimura, The involvement of aldose reductase in diabetic complications, Diabetes Metab. Rev. 4 (1988) 323 /337. [2] A.Y. Lee, S.K. Chung, S.S. Chung, Demonstration that polyol accumulation is responsible for diabetic cataract by the use of transgenic mice expressing the aldose reductase gene in the lens, Proc. Natl. Acad. Sci. USA 92 (1995) 2780 /2784. [3] L. Costantino, G. Rastelli, P. Vianello, G. Cignarella, D. Barlocco, Diabetes complications and their potential prevention: aldose reductase inhibition and other approaches, Med. Res. Rev. 19 (1999) 3 /23. [4] P.J. Oates, B.L. Mylari, Aldose reductase inhibitors: therapeutic implications for diabetic complications, Expert Opin. Invest. Drugs 8 (1999) 2095 /2119. [5] The Diabetes Control and Complications Trial Research Group, The effect of intensive diabetes treatment on the progression of diabetic retinopathy in insulin-dependent diabetes mellitus. The Diabetes Control and Complications Trial, Arch. Ophthal. 113 (1995) 36 /51. [6] C.E. Grimshaw, E.J. Mathur, Immunoquantitation of aldose reductase in human tissues, Anal. Biochem. 176 (1989) 66 /71.
332
Q. Chang et al. / Chemico-Biological Interactions 143 /144 (2003) 325 /332
[7] D.J. Hyndman, T.G. Flynn, Sequence and expression levels in human tissues of a new member of the aldo / keto reductase family, Biochim. Biophys. Acta 1399 (1998) 198 /202. [8] D. Cao, S.T. Fan, S.S. Chung, Identification and characterization of a novel human aldose reductase-like gene, J. Biol. Chem. 273 (1998) 11429 /11435. [9] S. Srivastava, A. Chandra, N.H. Ansari, S.K. Srivastava, A. Bhatnagar, Identification of cardiac oxidoreductase(s) involved in the metabolism of the lipid peroxidationderived aldehyde-4-hydroxynonenal, Biochem. J. 329 (1998) 469 /475. [10] S. Srivastava, T.M. Harter, A. Chandra, A. Bhatnagar, S.K. Srivastava, J.M. Petrash, Kinetic studies of FR-1, a growth factor-inducible aldo /keto reductase, Biochemistry 37 (1998) 12909 /12917. [11] D.L. Vander Jagt, B. Robinson, K.K. Taylor, L.A. Hunsaker, Reduction of trioses by NADPH-dependent aldo /keto reductases, aldose reductase, methylglyoxal, and diabetic complications, J. Biol. Chem. 267 (1992) 4364 /4369. [12] S. Srivastava, A. Chandra, L.F. Wang, W.E.J. Seifert, B.B. DaGue, N.H. Ansari, S.K. Srivastava, A. Bhatnagar, Metabolism of the lipid peroxidation product, 4-hydroxy-trans -2-nonenal, in isolated perfused rat heart, J. Biol. Chem. 273 (1998) 10893 /10900. [13] H.T. Ho, S.K. Chung, J.W. Law, B.C. Ko, S.C. Tam, H.L. Brooks, M.A. Knepper, S.S. Chung, Aldose reductasedeficient mice develop nephrogenic diabetes insipidus 10, Mol. Cell. Biol. 20 (2000) 5840 /5846. [14] I. Tarle, D.W. Borhani, D.K. Wilson, F.A. Quiocho, J.M. Petrash, Probing the active site of human aldose reductase.
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
Site-directed mutagenesis of Asp-43, Tyr-48, Lys-77, and His-110, J. Biol. Chem. 268 (1993) 25687 /25693. A. Kuhn, C. van Zyl, A. van Tonder, B.A. Prior, Purification and partial characterization of an aldo /keto reductase from Saccharomyces cerevisiae , Appl. Environ. Microbiol. 61 (1995) 1580 /1585. K. Nakamura, S. Kondo, Y. Kawai, N. Nakajima, A. Ohno, Amino acid sequence and characterization of aldo / keto reductase from bakers’ yeast, Biosci. Biotechnol. Biochem. 61 (1997) 375 /377. S.T. Kim, W.K. Huh, B.H. Lee, S.O. Kang, D-arabinose dehydrogenase and its gene from Saccharomyces cerevisiae , Biochim. Biophys. Acta 1429 (1998) 29 /39. A. Garay-Arroyo, A.A. Covarrubias, Three genes whose expression is induced by stress in Saccharomyces cerevisiae , Yeast 15 (1999) 879 /892. U. Oechsner, V. Magdolen, W. Bandlow, A nuclear yeast gene (GCY) encodes a polypeptide with high homology to a vertebrate eye lens protein, FEBS Lett. 238 (1988) 123 / 128. J.M. Petrash, B.S. Murthy, M. Young, K. Morris, L. Rikimaru, T.A. Griest, T. Harter, Functional genomic studies of aldo /keto reductases, Chem. Biol. Interact. 130 /132 (2001) 673 /683. E. Hur, D.K. Wilson, Crystallization and aldo /keto reductase activity of Gcy1p from Saccharomyces cerevisiae , Acta Crystallogr. D 56 (2000) 763 /765. H.C. Causton, B. Ren, S.S. Koh, C.T. Harbison, E. Kanin, E.G. Jennings, T.I. Lee, H.L. True, E.S. Lander, R.A. Young, Remodeling of yeast genome expression in response to environmental changes, Mol. Biol. Cell 12 (2001) 323 /337.
Aldo keto reductases as modulators of stress response /
Qing Chang, Theresa M. Harter, Loryn T. Rikimaru, J. Mark Petrash * Departments of Ophthalmology and Visual Sciences and of Genetics, Washington University School of Medicine, 660 S. Euclid Avenue, Campus Box 8096, St. Louis, MO 63110, USA
Abstract Human aldose reductase (AKR1B1) has been implicated as a factor in the pathogenesis of diabetic complications. However, little is known about the physiological role of this enzyme or of related aldo /keto reductases in human tissues. In mammalian systems, a gene knock out approach is often employed as an experimental strategy to probe for gene function. However, in the murine system, phenotypic characterization of an aldose reductase (AKR1B3) knock out is likely to be complicated due to functional compensation by redundant AKRs including AKRs 1A (aldehyde reductase), 1B7 (FR-1) and 1B8 (MVDP). As an alternate strategy, we are examining the budding yeast Saccharomyces cerevisiae as a model system for a functional genomics study of AKRs. A distinct advantage of this system centers on the ability to readily ablate multiple targeted genes in a single strain. In addition to providing insights into functional redundancy, this system allows us to use a genetic approach to study possible effector pathways associated with one or more individual genes. Yeast open reading frames (ORFs) encoding AKRs with functional similarity to human aldose reductase (AKR1B1) were identified by BLAST analysis and were functionally validated by studies of recombinant proteins. By ablating three of the yeast AKR genes most functionally similar to AKR1B1, we have created a unique strain of S. cerevisiae that shows enhanced sensitivity to stress. Ongoing studies with oligonucleotide arrays show that the triple null strain has an altered transcription profile consistent with an enhanced stress response in comparison with the parental strain. These data indicate that AKR-null strains may provide new insights into signaling mechanisms involving this family of proteins. # 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Aldo /keto reductase; Saccharomyces cerevisiae ; Gene chip; Stress response; Suppressor
1. Introduction Diabetes mellitus is recognized as a leading cause of new cases of blindness among Americans between the ages of 20 and 74 years. In comparison with the general population, diabetic patients
* Corresponding author. Tel.: /1 (314) 362-3335; fax: /1 (314) 362-3638 E-mail address: [email protected] (J.M. Petrash).
face a 25-fold increased risk of blindness resulting from diabetic retinopathy or cataract. While many theories have been advanced to explain mechanisms leading to diabetic complications, a prominent model suggests that enhanced flux of glucose through the polyol pathway may be causally linked to diabetic complications. Two enzymes make up the polyol pathway: aldose reductase (AKR1B) and sorbitol dehydrogenase, a short chain dehydrogenase/reductase. Evidence linking enhanced polyol pathway activity and diabetic
0009-2797/02/$ - see front matter # 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 0 0 9 - 2 7 9 7 ( 0 2 ) 0 0 2 1 6 - 8
326
Q. Chang et al. / Chemico-Biological Interactions 143 /144 (2003) 325 /332
complications comes mostly from animal studies where severity of disease can be altered by pharmacologic blockade [1] or genetic enhancement [2] of aldose reductase activity. Despite success in preventing complications in experimentally diabetic animal models treated with aldose reductase inhibitors, translation of this strategy to human therapy has not been successful [3]. Various explanations have been advanced to rationalize this discordance [4]. Perhaps of prime importance, the Diabetes Control and Complications Trial provided compelling evidence that early intervention is an important element in successfully arresting or preventing the onset and/or progression of diabetic complications [5]. Despite the substantial investment of human and economic capital in the study of aldose reductase inhibition, very little is known about the physiological role of this enzyme. The most abundant tissue source of aldose reductase is the adrenal gland, but its physiological role in that specialized tissue is not known [6]. Human tissues contain multiple aldo /keto reductases with structural and catalytic properties strikingly similar to aldose reductase. Newly discovered aldose reductase-like proteins include the human small intestine reductase AKR1B10 [7] (also known as human aldose reductase-like protein 1 [8]). Kinetic studies of aldose reductase and related AKRs point to a possible role for these enzymes in aldehyde detoxification [9 /11]. However, functional proof that aldose reductase fulfills this role in vivo has been limited to the ischemia /reperfusion process in the cardiovascular system [12]. Ablation of a targeted gene can often provide useful insight into possible physiological roles, but functional redundancy among similar aldo /keto reductases complicates this strategy in available animal models. Indeed, knock out mice resulting from ablation of the mouse aldose reductase gene were without a meaningful phenotype that could be related to aldose reductase function outside the kidney [13]. As an alternative to gene ablation in the murine animal model, we began an investigation into use of Saccharomyces cerevisiae as a model system.
1.1. Multiplicity of yeast aldo /keto reductases Six yeast open reading frames (ORFs) that encode polypeptides with sequence similarity to human aldose reductase (AKR1B1) can be identified by BLAST analysis. In Fig. 1, the encoded polypeptides are shown aligned with the AKR1B1 sequence. From the sequence alignment, it can be seen that all of the putative yeast AKRs are of a molecular mass very similar to the human enzyme. Further, it can be appreciated that all have conserved those amino acid side chains that were previously demonstrated to be of key importance in the catalytic activity of mammalian AKRs, including aldose reductase (AKR1B1), aldehyde reductase (AKR1A), and several of the hydroxysteroid dehydrogenases (e.g. AKR 1C1) [14]. These side chains include Asp-43, Tyr-48, Lys-88 and His-110 (numbered according to the human aldose reductase primary sequence). Previous studies have shown that some but not all of the proteins encoded by the putative yeast AKR ORFs can function as active enzymes. Table 1 shows the AKR ORFs, their conventional AKR designations, and any associated functional nomenclature from the published literature. Proteins derived from the six yeast AKRs can be sorted into three subfamilies of the AKR protein superfamily (see http://www.med.upenn.edu/akr/): YHR104Wp (AKR2B6) is categorized into subfamily 2 along with xylose reductases from other yeast species. Polypeptides encoded by three ORFs can be sorted into AKR subfamily 3. These include YOR120Wp (AKR3A1), YDR368Wp (AKR3A2) and YBR149Wp (AKR3C). Subfamily 5 contains two of the putative yeast AKRs, including YJR096Wp (AKR5F) and YDL124Wp (AKR5G). Interestingly, proteins from subfamilies 2 and 3 (YHR104W, YDR368W, and YBR149W) were initially identified on the basis of their enzymatic functions [15 /17]. Thus, YBR149W encodes one subunit of the heterodimeric enzyme arabinose dehydrogenase gene (ARA1 ) [17], while YDR368W is synonymous with the YPR1 gene, which encodes a xylose reductase [15,16]. The gene encoding YHR104Wp was discovered on the basis of its induction in response to stress and was named GRE3 (genes de respuestra a estres) [18];
Q. Chang et al. / Chemico-Biological Interactions 143 /144 (2003) 325 /332
327
Fig. 1. Predicted sequences of yeast aldo /keto reductases. Yeast sequences are identified by their seven character ORF names. Alignments with the primary sequence of human aldose reductase (HAR) were edited using GeneDoc (http://www.psc.edu/biomed/ genedoc/). Amino acids shown in previous studies to play catalytic roles in human aldo /keto reductases are identified by shading.
additional studies demonstrated aldo /keto reductase activity associated with YHR104Wp [15]. While initial discovery of GCY1 (galactose-inducible crystallin-like protein) centered on its induc-
tion by galactose [19], we and others subsequently demonstrated that Gcy1p is a relatively efficient aldo /keto reductase [20,21]. To date, no studies have implicated a functional role for YJR096Wp,
328
Q. Chang et al. / Chemico-Biological Interactions 143 /144 (2003) 325 /332
Table 1 Aldo /keto reductases in S. cerevisiae ORF designation AKR designation Gene name Reported function
YHR104W AKR2B6 GRE3 Aldo /keto reductase
YOR120W AKR3A1 GCY1 Aldo /keto reductase
YDR368W AKR3A2 YPR1 Xylose reductase
YBR149W AKR3C ARA1 Arabinose dehydrogenase
YJR096W AKR5F
YDL124W AKR5G
although we previously demonstrated this protein to have aldo /keto reductase activity with a variety of substrates [20]. As we reported previously, the yeast system provides many advantages for functional characterization of genes comprising a complex gene family [20]. Since our studies ultimately seek to establish a physiological role for human aldose reductase, we hypothesized that the sequence of yeast AKRs selected for evaluation should be determined by the degree of structural and catalytic similarity to the human enzyme. Based in part on kinetic and structural considerations summarized in our previously published work [20], we have chosen to evaluate the yeast AKR genes in the following order of priority: YPR1, GRE3, GCY1, YJR096W, YDL124W, ARA1. Work presented in the present paper focuses exclusively on YPR1, GRE3 and GCY1.
deletion strains would be phenotypically indistinguishable from wild type due to functional compensation by remaining AKR genes. Indeed, in phenotypic screens for temperature dependence on growth, carbon auxotrophy and morphology, the deletion strains comprising either one or two AKR gene disruptions were indistinguishable from the parental strain. In contrast, the triple AKR null strain ypr1 Dgre3 Dgcy1 D showed a marked difference from the wild type strain in heat shock sensitivity (Fig. 2). It is interesting to note that the threshold for emergence of a clear phenotype appears to lie at three AKR genes, as deletion of two AKR genes, in any combination, fails to give rise to a measurable phenotype, whereas deletion of three AKR genes gives rise to a heat shock phenotype. The presence of any one of the three catalytically similar yeast AKRs appears to be sufficient to confer protection against heat shock sensitivity.
2. Results and discussion
2.2. Gene chip studies
2.1. Gene disruption studies
As an additional strategy to characterize the triple null AKR mutant, we carried out transcription profiling studies with the use of oligonucleotide arrays. To ascertain changes in gene expression under basal and heat shock conditions, total RNA was extracted from triplicate wild type and triple AKR null cultures. For heat shock treatment, mid-log phase cultures at 30 8C were shifted to 37 8C for 30 min. Cultures under basal conditions were maintained at 30 8C. Oligonucleotide array chips were run in triplicate for each condition and for each yeast strain, resulting in a total of 12 gene chips in this analysis. Array data were analyzed using the AFFYMETRIX MICROARRAY SUITE (v5.0) of programs and GENESPRING (Silicon Genetics) for filtering and pathway mapping. We established a cutoff of 2.5-fold as the
A series of isogenic yeast strains was constructed to represent all possible permutations of yeast AKR gene deletions involving YPR1, GRE3 and GCY1. Thus, strains were constructed with a single AKR gene deletion (ypr1 D, gre3 D, gcy1 D), as well as strains with two AKR genes deleted (ypr1 D gre3 D; ypr1 Dgcy1 D; gre3 Dgcy1 D). Finally, a triple AKR null strain was developed (ypr1 Dgre3 Dgcy1 D). Each strain was analyzed for phenotypic changes, utilizing a battery of tests selected to screen for changes in temperature sensitivity, carbon auxotrophy, morphology and stress response. Given the broad overlap in functional similarity, at least at the level of aldehyde reduction, we predicted that single AKR gene
Q. Chang et al. / Chemico-Biological Interactions 143 /144 (2003) 325 /332
329
Fig. 2. Heat shock sensitivity of triple AKR null strain. Serially-diluted cultures grown under control conditions or following heat shock treatment (55 8C, 2 min) were spotted on agar plates for growth at 30 8C. In comparison with the wild type strain, cultures of the triple AKR null strain are substantially diminished following heat shock treatment. Fig. 3. Venn diagram of genes responding to heat shock. Wild type cells responded with 448 genes showing ]/2.5-fold increase in transcripts (red and purple) while the triple AKR null strain responded with 61 genes showing a ]/2.5-fold increase in transcripts (blue and purple). A comparison of the genes identified from the two strains showed that 50 were in common (purple). Fig. 4. Venn diagram of wild type and triple null strains. Wild type cells responded with 448 genes showing ]/2.5-fold increase in transcripts (red and yellow). Comparison of the triple AKR null strain with wild type grown under basal conditions shows that the triple null strain has 241 genes that are constitutively upregulated by a factor of ]/2.5-fold (green and yellow). A comparison of the genes identified from the two growth conditions shows that 143 were in common (yellow).
threshold to define what we consider to be a meaningful change in transcript level. Examination of wild type cells revealed that heat shock resulted in the induction of 448 genes (Fig. 3). Similar examination of the triple AKR null cells showed that only 61 genes were induced by heat shock treatment. A comparison of the induced genes identified in the WT and triple AKR
null strain showed that most of the heat shock induced genes in the triple null strain were also induced by heat shock in the WT strain (Fig. 3). This demonstrates that the capability to mount a heat shock response is largely preserved in the triple AKR null strain. A comparison of WT and triple null strains cultured under basal conditions revealed that 241
330
Q. Chang et al. / Chemico-Biological Interactions 143 /144 (2003) 325 /332
Fig. 5. Venn diagram of heat shock response in wild type and triple null strains. Area of circle marked by RED represents induced genes in wild type cells subjected to heat shock. Area of circle marked by GREEN represents induced genes in triple AKR null cells grown under basal conditions. Area of circle marked by BLUE represents induced genes in triple AKR null cells subjected to heat shock. The area representing overlap of all three circles shows that 15 genes in common were upregulated in all comparison sets. Fig. 6. Functional categories represented by constitutively upregulated genes in the triple AKR null strain.
Q. Chang et al. / Chemico-Biological Interactions 143 /144 (2003) 325 /332
331
Table 2 Specificity of transcriptional remodeling Category
# Genes induced
Total # ORFS in category
Percentage induced
Energy/metabolism Gene expression Cellular organization Transport Signaling/interactions Cell cycle/DNA processing Unclassified Total
36 34 47 14 7 30 43 215
1318 1725 2258 808 467 906 2399 10 556
2.7 2.0 2.1 1.7 1.5 3.3 1.8 2.0
genes were induced at least 2.5-fold in the triple null strain. As shown in the Venn diagram in Fig. 4, approximately 60% of the constitutively-induced genes in the triple AKR null strain (143 of 241) were identified as heat shock responsive genes in the WT. Indeed, of the 143 constitutively upregulated genes in the triple null strain that were heat shock induced in the WT, only 15 (10%) were further upregulated by heat shock in the triple AKR null strain (Fig. 5). With the use of pathway mapping tools, we sorted each of the constitutively upregulated genes into functional categories. Fig. 6 shows that upregulated genes could be sorted into major categories involved in metabolism, gene expression, cellular organization and transport. Table 2 shows that constitutively induced genes in the triple AKR null strain comprise approximately 2% of all genes in the sample pool. Thus, it appears that the transcriptome of the AKR triple null strain reflects constitutive activation of many stress response genes. This suggests a potential role for aldo /keto reductases in downregulation of genes that are controlled by the stress response pathway. Since many stress response genes are activated through the action of Msn2/ Msn4 transcriptional factors [22], we hypothesize that the aldo /keto reductases identified in the present study could act as signaling modulators along the pathway leading to regulation of Msn2p and/or Msn4p. In this context, lack of sufficient AKR functionality could result in constitutive activation of Msn2p/Msn4p, leading to dysregulated transcription of stress response genes.
Acknowledgements This work was supported in part from NIH grants EY05856, EY02687, and 2P60DK20579. Additional support was provided by Research to Prevent Blindness, Inc. through a Lew R. Wasserman Award (J.M. Petrash) and an unrestricted award to the Washington University Department of Ophthalmology and Visual Sciences. The support of the Kilo Diabetes and Vascular Research Foundation is also gratefully acknowledged.
References [1] J.H. Kinoshita, C. Nishimura, The involvement of aldose reductase in diabetic complications, Diabetes Metab. Rev. 4 (1988) 323 /337. [2] A.Y. Lee, S.K. Chung, S.S. Chung, Demonstration that polyol accumulation is responsible for diabetic cataract by the use of transgenic mice expressing the aldose reductase gene in the lens, Proc. Natl. Acad. Sci. USA 92 (1995) 2780 /2784. [3] L. Costantino, G. Rastelli, P. Vianello, G. Cignarella, D. Barlocco, Diabetes complications and their potential prevention: aldose reductase inhibition and other approaches, Med. Res. Rev. 19 (1999) 3 /23. [4] P.J. Oates, B.L. Mylari, Aldose reductase inhibitors: therapeutic implications for diabetic complications, Expert Opin. Invest. Drugs 8 (1999) 2095 /2119. [5] The Diabetes Control and Complications Trial Research Group, The effect of intensive diabetes treatment on the progression of diabetic retinopathy in insulin-dependent diabetes mellitus. The Diabetes Control and Complications Trial, Arch. Ophthal. 113 (1995) 36 /51. [6] C.E. Grimshaw, E.J. Mathur, Immunoquantitation of aldose reductase in human tissues, Anal. Biochem. 176 (1989) 66 /71.
332
Q. Chang et al. / Chemico-Biological Interactions 143 /144 (2003) 325 /332
[7] D.J. Hyndman, T.G. Flynn, Sequence and expression levels in human tissues of a new member of the aldo / keto reductase family, Biochim. Biophys. Acta 1399 (1998) 198 /202. [8] D. Cao, S.T. Fan, S.S. Chung, Identification and characterization of a novel human aldose reductase-like gene, J. Biol. Chem. 273 (1998) 11429 /11435. [9] S. Srivastava, A. Chandra, N.H. Ansari, S.K. Srivastava, A. Bhatnagar, Identification of cardiac oxidoreductase(s) involved in the metabolism of the lipid peroxidationderived aldehyde-4-hydroxynonenal, Biochem. J. 329 (1998) 469 /475. [10] S. Srivastava, T.M. Harter, A. Chandra, A. Bhatnagar, S.K. Srivastava, J.M. Petrash, Kinetic studies of FR-1, a growth factor-inducible aldo /keto reductase, Biochemistry 37 (1998) 12909 /12917. [11] D.L. Vander Jagt, B. Robinson, K.K. Taylor, L.A. Hunsaker, Reduction of trioses by NADPH-dependent aldo /keto reductases, aldose reductase, methylglyoxal, and diabetic complications, J. Biol. Chem. 267 (1992) 4364 /4369. [12] S. Srivastava, A. Chandra, L.F. Wang, W.E.J. Seifert, B.B. DaGue, N.H. Ansari, S.K. Srivastava, A. Bhatnagar, Metabolism of the lipid peroxidation product, 4-hydroxy-trans -2-nonenal, in isolated perfused rat heart, J. Biol. Chem. 273 (1998) 10893 /10900. [13] H.T. Ho, S.K. Chung, J.W. Law, B.C. Ko, S.C. Tam, H.L. Brooks, M.A. Knepper, S.S. Chung, Aldose reductasedeficient mice develop nephrogenic diabetes insipidus 10, Mol. Cell. Biol. 20 (2000) 5840 /5846. [14] I. Tarle, D.W. Borhani, D.K. Wilson, F.A. Quiocho, J.M. Petrash, Probing the active site of human aldose reductase.
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
Site-directed mutagenesis of Asp-43, Tyr-48, Lys-77, and His-110, J. Biol. Chem. 268 (1993) 25687 /25693. A. Kuhn, C. van Zyl, A. van Tonder, B.A. Prior, Purification and partial characterization of an aldo /keto reductase from Saccharomyces cerevisiae , Appl. Environ. Microbiol. 61 (1995) 1580 /1585. K. Nakamura, S. Kondo, Y. Kawai, N. Nakajima, A. Ohno, Amino acid sequence and characterization of aldo / keto reductase from bakers’ yeast, Biosci. Biotechnol. Biochem. 61 (1997) 375 /377. S.T. Kim, W.K. Huh, B.H. Lee, S.O. Kang, D-arabinose dehydrogenase and its gene from Saccharomyces cerevisiae , Biochim. Biophys. Acta 1429 (1998) 29 /39. A. Garay-Arroyo, A.A. Covarrubias, Three genes whose expression is induced by stress in Saccharomyces cerevisiae , Yeast 15 (1999) 879 /892. U. Oechsner, V. Magdolen, W. Bandlow, A nuclear yeast gene (GCY) encodes a polypeptide with high homology to a vertebrate eye lens protein, FEBS Lett. 238 (1988) 123 / 128. J.M. Petrash, B.S. Murthy, M. Young, K. Morris, L. Rikimaru, T.A. Griest, T. Harter, Functional genomic studies of aldo /keto reductases, Chem. Biol. Interact. 130 /132 (2001) 673 /683. E. Hur, D.K. Wilson, Crystallization and aldo /keto reductase activity of Gcy1p from Saccharomyces cerevisiae , Acta Crystallogr. D 56 (2000) 763 /765. H.C. Causton, B. Ren, S.S. Koh, C.T. Harbison, E. Kanin, E.G. Jennings, T.I. Lee, H.L. True, E.S. Lander, R.A. Young, Remodeling of yeast genome expression in response to environmental changes, Mol. Biol. Cell 12 (2001) 323 /337.