Aldo–keto reductase family 1 B10 gene silencing results in growth inhibition of colorectal cancer cells: Implication for cancer intervention

Abstract

Aldo–keto reductase family 1 B10 (AKR1B10), a member of aldo–keto reductase superfamily, is overexpressed in human hepatocellular carcinoma, lung squamous cell carcinoma and lung adenocarcinoma. Our previous study had demonstrated that the ectopic expression of AKR1B10 in 293T cells promotes cell proliferation. To evaluate its potential as a target for cancer intervention, in the current study we knocked down AKR1B10 expression in HCT-8 cells derived from a colorectal carcinoma, using chemically synthesized small interfering RNA (siRNA). The siRNA 1, targeted to encoding region, downregulated AKR1B10 expression by more than 60%, and siRNA 2, targeted to 3′ untranslational region, reduced AKR1B10 expression by more than 95%. AKR1B10 silencing resulted in approximately a 50% decrease in cell growth rate and nearly 40% suppression of DNA synthesis. More importantly, AKR1B10 downregulation significantly reduced focus formation rate and colony size in semisolid culture, indicating the critical role of AKR1B10 in HCT-8 cell proliferation. Recombinant AKR1B10 protein showed strong enzymatic activity to acrolein and crotonaldehyde, with K_m = 110.1 ± 12.2 μM and V_max = 3,122.0 ± 64.7 nmol/mg protein/min for acrolein and K_m = 86.7 ± 14.3 μM and V_max = 2,647.5 ± 132.2 nmol/mg protein/min for crotonaldehyde. AKR1B10 downregulation enhanced the susceptibility of HCT-8 cells to acrolein (25 μM) and crotonaldehyde (50 μM), resulting in rapid oncotic cell death characterized with lactate dehydrogenase efflux and annexin-V staining. These results suggest that AKR1B10 may regulate cell proliferation and cellular response to additional carbonyl stress, thus being a potential target for cancer intervention. © 2007 Wiley-Liss, Inc.

Aldo–keto reductase family 1 B10 (AKR1B10, also designated aldose reductase-like-1, ARL-1) is a novel member of aldo–keto reductase (AKR) superfamily, isolated from human hepatocellular carcinoma (HCC).1 The AKR superfamily is involved in intracellular detoxification, carcinogenesis and cancer therapeutics.2, 3 Enhanced expression of aldose reductase (AR, also referred to as AKR1B1) is recognized in many types of tumors.1, 4, 5 Inhibition of AR activity results in cancer cell growth inhibition and susceptibility to carbonyl compounds and chemotherapeutic agents,6, 7 while induction of AR expression leads to tumor cell resistance to anticancer drugs.8 Therefore, AR inhibitors developed for the treatment of diabetic complications have become potential chemotherapeutic agents for cancers with AR overexpression.7, 9 AR also regulates mitogenic signaling pathways in vascular smooth muscle cells and vascular endothelial cells, controlling cells growth.10, 11

AKR1B10 shows 71% amino acid sequence identity to AR.1 Unlike the ubiquitous expression of AR, the AKR1B10 gene is primarily expressed in the small intestine and colon, with lower levels in the liver, thymus, prostate and testes.1AKR1B10 is overexpressed in ∼54% of the HCC tissues and in 84.4% of lung squamous cell carcinomas (SCC) and 29.2% of lung adenocarcinomas in smokers, suggesting its potential as a tumor marker.1, 12, 13 However, little is known about its role in cell carcinogenesis and cancer interventions in these tumor types. A recent study in our laboratory has demonstrated that transient delivery of AKR1B10 into 293T cells significantly prevented the cytotoxicity of acrolein, a highly reactive carbonyl compound.14

Reactive carbonyls, such as acrolein and crotonaldehyde, are constantly produced in living cells via lipid peroxidation in addition to their wide existence in living environments and various diets.15-17 These carbonyls can interact with free amino and sulfhydryl groups of proteins, peptides and amino acids, forming covalently modified adducts.18, 19 These nonspecific, covalent modifications may cause protein dysfunction, resistance to intracellular proteolysis or depolymerization. Electrophilic carbonyls can also react with DNA nucleophiles, forming alkylated DNA adducts that block DNA replication, arrest transcription and cause DNA mutations and breaks.17, 20-23 Therefore, reactive carbonyls are cytotoxins that need to be quickly deactivated. In the current study, we investigated the effect of small interfering RNA (siRNA)-mediated downregulation of AKR1B10 on proliferation of colorectal cancer cells (HCT-8) and evaluated its potential as a tumor intervention target.

AKR1B10, aldo–keto reductase family 1 B10; ARL-1, aldose reductase-like-1; GI, gastrointestinal; LDH, lactate dehydrogenase; MDA, malondialdehyde; NADPH, Nicotinamide adenine dinucleotide phosphate (reduced); PMSF, Phenylmethanesulfonyl fluoride; PS, phospholipid phosphatidylserine; and siRNA, small interfering RNA.

Material and methods

Cell culture

HCT-8 cells, purchased from American Type Culture Collection (Manassas, VA), were maintained in RPMI-1640 medium (Hyclone, UT) containing 10% fetal bovine serum (FBS), 2 mM glutamine, 100 U/ml penicillin and 100 μg/ml streptomycin at 37°C, 5% CO₂.

AKR1B10 and AR protein preparation and enzymatic activity assays

AKR1B10 and AR recombinant proteins were prepared as described previously.1 Acrolein and crotonaldehyde substrate activity was determined in a mixture consisting of 135 mM sodium phosphate (pH 7.0), 0.2 mM NADPH, 1.0 mM β-mercaptoethanol, 50 mM KCl, 2 μg AKR1B10 or AR protein and appropriate substrates, ranging from 0 to 20 mM. Enzymatic reactions were conducted at 35°C for 20 min. Oxidized NADPH was measured at OD₃₄₀ to express enzymatic activity.1 Reaction mixtures without AKR1B10 or AR proteins were used as blank controls. Michaelis–Menten constants (K_m and V_max) were calculated via Lineweaver–Burk plots using GraphPad Prism 4 (Graph Pad Software, CA). To measure AKR1B10 activity in HCT-8 cells, cells were lysed on ice in a buffer containing 20 mM NaH₂PO₄, 2 mM DTT, 5 μM leupeptin and 20 μM PMFS for 30 min, followed by a centrifugation at 10,000g, 4°C for 10 min. Soluble proteins (50 μg) were used for AKR1B10 activity assay, using 20 mM DL-glyceraldehyde as a substrate.1 All experiments were repeated 3 times.

AKR1B10 silencing by siRNAs

Two siRNAs targeted to encoding (siRNA 1, 5′GCAAGUUGUGGCCCACUUUtt) and 3′ untranslational (siRNA 2, 5′CGAGAAUCGAGGUGCUGUUtt) regions of AKR1B10 were chemically synthesized (Ambion, TX). A scrambled siRNA was used as a negative control. For siRNA delivery, HCT-8 cells (3.5 × 10^4–5 in Opti-MEM I medium) were mixed gently with siRNA and OligofectAMINE (Invitrogen, CA) in a volume of 0.5–1.5 ml, following manufacturer's instructions, and incubated at 37°C, 5% CO₂ for 4 hr, followed by the addition of an equal volume of fresh medium containing 20% FBS. Cells were continuously incubated until harvest.

Western blot analysis

Western blot was performed as previously described using 30 μg of soluble proteins extracted from HCT-8 cells.14 Protein amounts were corrected by β-actin monoclonal antibody (1:40,000, Sigma, MO).

Cell growth, ³H-thymidine incorporation and clonogenic assay

For cell growth tests, 1 × 10⁴ of HCT-8 cells/well were seeded into 24-well plates and incubated at 37°C, 5% CO₂. At indicated time points, the viable cells were trypsinized and counted with trypan blue exclusion. Three wells at each time point were used to obtain the averages and 3 independent repeats were performed for statistical tests.

³H-thymidine incorporation into DNA was determined by pulsing 1 × 10⁴ cells/well in 24-well plates with 10 μCi/well of ³H-thymidine (MP Biochemicals, OH) for 2 hr. Cells were washed with PBS and lysed in 0.4 ml of 15% trichloroacetic acid (TCA) on ice. After washing twice with 15% TCA, acidic-insoluble materials were dissolved in 50 μl of 0.1 N NaOH. An aliquot (10 μl) was used to determine protein amounts using protein assay reagents (Bio-Rad, CA) and the remaining was subjected to radioactivity measurements. ³H-thymidine incorporation into DNA was expressed by cpm/μg protein. The experiments were repeated 3 times with 3 wells for each treatment.

Clonogenic growth of HCT-8 cells was evaluated by anchorage-independent growth in soft agar. In 24-well plates, 100 cells/well were suspended in 0.5 ml of 0.3% Noble agar (Sigma, MO) and layered over 0.5 ml of 0.5% agar in the same medium. Three wells were prepared for each treatment. After being cultured at 37°C, 5% CO₂ for 2 weeks, foci were photographed and scored under inverted microscope. Clonogenic efficiency (%) was calculated as: (number of clones/number of seeded cells) × 100. The experiments were repeated 3 times.

Acrolein and crotonaldehyde toxicity and flow cytometry analysis

Cytotoxicity of acrolein and crotonaldehyde was tested by exposing HCT-8 cells (3 × 10⁵/well) in 12-well plates to a single dose of acrolein (25 μM) or crotonaldehyde (50 μM) in full medium for 72 hr. After trypsinization, viable cells were counted by trypan blue exclusion. Three independent experiments were performed with 3 wells for each treatment.

For flow cytometry analysis, 6 × 10⁵ of HCT-8 cells in 60-mm cell culture dishes were exposed to 25-μM acrolein in medium for 24 hr. Medium was gently removed, and cells were washed with cold PBS and then trypsinized. Cells in PBS and trypsin digestion were pooled, washed with PBS twice at 1,200g for 10 min and then subjected to immediate propidium iodide (PI) and annexin-V-FITC staining for 10 min in the dark. FACScan analysis was performed using a FACScan cytometer (Becton Dickinson, CA). Three repeats were conducted.

Lactate dehydrogenase leakage assay

Lactate dehydrogenase (LDH) efflux assay was performed with LDH assay kit (Roche, IN). HCT-8 cells (1 × 10⁴ cells/well) in 24-well plates were treated with 25 μM of acrolein in full medium for 12 or 24 hr. Medium was collected and cells were lysed for 10 min in 0.5% (v/v) Triton X-100 in 0.1 M potassium phosphate buffer (pH 7.4). Supernatant was removed for LDH assays after a centrifugation at 10,000g for 5 min. LDH activity was measured following manufacturer's instructions. LDH leakage was calculated as: LDH release (%) = [LDH in medium/(LDH in medium + LDH in cell lysates)] × 100. The experiments were independently performed 3 times with 3 wells each treatment.

Statistic analysis

Statistic analysis was performed using Student's t test with INSTAT statistical analysis package (Graph Pad Software, CA). Significance was defined as p < 0.05.

Results

Effect of siRNA-mediated silencing of AKR1B10 on cancer cell proliferation and clonogenic growth

To understand its intracellular function, AKR1B10 was silenced by chemically synthesized siRNAs. Human colorectal cancer cells (HCT-8) with abundant AKR1B10 expression were used for the study of AKR1B10 silencing. AR expression is undetectable in this cell line (data not shown). Two siRNAs, targeted to encoding (siRNA 1) and 3′ untranslational (siRNA 2) regions of AKR1B10 mRNA were used to confirm the specificity of the AKR1B10 knockdown. A scrambled siRNA was used as a negative control. As demonstrated in complementary data (Fig. 1), both siRNA 1- and siRNA 2-induced specific AKR1B10 downregulation in HCT-8 cells. The maximal silencing occurred during 48–96 hr after delivery and 50 nM of siRNA represented the optimal concentration for both siRNA 1 and 2.

Figure 1. Effect of AKR1B10 silencing on cell growth and DNA synthesis. HCT-8 cells were transfected with 50 nM siRNA and incubated in 6-well plates for Western blot (a), enzymatic activity (Unit: nmoles/mg protein/hour) (b) and [³H]-thymidine incorporation (cpm/μg protein) (d). A portion of transfected cells were spread into 24-well plates at 10⁴ cells/well for cell growth assays (c). Western blot, enzymatic activity, and [³H]-thymidine incorporation were conducted at 72 hr after transfection to allow for AKR1B10 knockdown. Viable cells were counted at indicated time points. All values represent mean ± SD from 3 independent experiments. *, p < 0.05. Oligo, oligofectAMINE (transfection reagent); and Scrambled, scrambled siRNA.

Growth rate and DNA synthesis of HCT-8 cells with AKR1B10 downregulation were examined. In the presence of 50 nM siRNAs, AKR1B10 protein levels (Fig. 1a) and enzymatic activity (Fig. 1b) in HCT-8 cells were both reduced by more than 60% (siRNA 1) and 95% (siRNA 2). In turn, this AKR1B10 knockdown resulted in significant growth inhibition and DNA synthesis suppression. In AKR1B10 downregulated cells, the viable cells were reduced by ∼50% at day 7 (Fig. 1c) and the incorporation of ³H-thymidine into DNA was decreased by nearly 40% (Fig. 1d), compared with the control cells without AKR1B10 silencing.

Clonogenic growth tests the viability and proliferation of individual cells and thus is a hallmark of cell carcinogenic transformation and an important feature of metastatic cancer cells. In this study, we further assessed the effect of AKR1B10 downregulation on the anchorage-independent growth of HCT-8 cells in soft agar. Figure 2 shows that AKR1B10 knockdown in HCT-8 cells resulted in significant decrease in focus formation and colony size, demonstrating reduced clonogenic growth.

Figure 2. Anchorage-independent growth. HCT-8 cells were transfected with siRNA (50 nM) and grown in soft agar as described in Material and Methods. (a) Clonogenic growth, ×10 objective. (b) Colony formation rate, defined as (colony number/seeded cell number) × 100. Values indicate the average ± SD from 3 independent repeats. *, p < 0.05. Oligo, oligofectAMINE; and Scrambled, scrambled siRNA.

Enzymatic activity of AKR1B10 protein and HCT-8 cell sensitivity to acrolein and crotonaldehyde

AKR1B10 protein is an enzyme catalyzing reduction of aromatic and aliphatic aldehydes.1 To understand the biochemical mechanisms of cell growth inhibition induced by AKR1B10 knockdown, we examined its in vitro enzymatic activity to cellular carbonyl by-products, acrolein and crotonaldehyde, and assessed the response of HCT-8 cells to these two compounds when AKR1B10 was silenced. AKR1B10 recombinant protein produced with a Qiagen prokaryotic expression system was used to measure its enzymatic activity and AR protein was prepared in parallel as a control. As shown in Table I, AKR1B10 demonstrated high affinity and strong activity toward both acrolein and crotonaldehyde, suggesting the potential role of AKR1B10 in regulating their intracellular concentrations. In the presence of acrolein (25 μM) and crotonaldehyde (50 μM), AKR1B10 downregulation resulted in dramatic HCT-8 cell death (Fig. 3) indicating the role of AKR1B10 in cellular response to additional carbonyl stress.

Figure 3. Acrolein and crotonaldehyde cytotoxicity. HCT-8 cells transfected with siRNAs (50 nM) were incubated for 48 hr to trigger AKR1B10 silencing, and then fed with fresh medium containing acrolein (25 μM) or crotonaldehyde (50 μM) (a single dose). After 72 hr, viable cells were counted by trypan blue exclusion. Values are the mean ± SD from 3 independent observations. Oligo, oligofectAMINE; Sc, scrambled siRNA; ACR, acrolein; Cr, crotonaldehyde; S1, siRNA 1; and S2, siRNA 2.

Table I. Kinetic Parameters of AR and AKR1B10 Proteins

Substrates	AKR1B10		AR
	Vmax [nmol/(mg min)]	Km (μM)	Vmax [nmol/(mg min)]	Km (μM)
AR and AKR1B10 protein preparation and enzymatic activity were performed as described in Material and Methods. Velocity: nmol/mg protein/min.
Acrolein	3,122.0 ± 64.7	110.1 ± 12.2	903.8 ± 103.5	30.3 ±16.8
Crotonaldehyde	2,647.5 ± 132.2	86.7 ± 14.3	1,216.2 ± 21.8	84.4 ± 8.5

Oncotic HCT-8 cell death induced by acrolein

Mechanisms of acrolein-induced HCT-8 cell death were further investigated by LDH efflux and FACScan analysis. In the presence of 25 μM of acrolein, LDH release into the medium was measured to examine the integrity of the membranes of dead cells; flow cytometry was used to check the staining of annexin-V-FITC, a Ca⁺⁺ dependent protein binding to phospholipid phosphatidylserine (PS) with high affinity.24 As shown in Figure 4, acrolein-induced HCT-8 cell death occurred with annexin-V-FITC staining, a characteristic of apoptotic or oncotic cells.

Figure 4. FACScan analysis of cell death. HCT-8 cells transfected with siRNAs (50 nM) were incubated for 48 hr to induce AKR1B10 downregulation, and then exposed to acrolein (25 μM) for 24 hr. FACScan analysis was conducted as described in the Material and Methods. (a) Cell distributions. (b) Annexin-V staining (%). Values represent mean ± SD from 3 independent measurements. PI, propidium iodide; Oligo, oligofectAMINE; Sc, scrambled siRNA; ACR, acrolein; S1, siRNA 1; and S2, siRNA 2.

LDH leakage into medium was 2- to 3-fold higher in HCT-8 cells with AKR1B10 knockdown than in scrambled siRNA control cells (Fig. 5), indicating that acrolein-induced HCT-8 cell death was accompanied with membrane disintegration. Combined with annexin-V staining data, it is apparent that in AKR1B10 downregulated HCT-8 cells, acrolein at greater concentrations induced oncotic cell death, as is the case with 293T cells.14

Figure 5. Lactate dehydrogenase (LDH) efflux. HCT-8 cells transfected with siRNAs (50 nM) were incubated for 48 hr to induce AKR1B10 silencing, and then exposed to acrolein (25 μM) for 12 or 24 hr. LDH activity in medium and in cell lysates was measured as described in the Materials and Methods. LDH leakage (%) = [LDH in medium/(LDH in medium + LDH in cell lysates)] × 100. Values are the average ± SD from 3 independent tests. Oligo, oligofectAMINE; and Scrambled, scrambled siRNA.

Discussion

Ectopic expression of the AKR1B10 gene in 293T cells allows for cell resistance to acrolein cytotoxicity.14 To explore its potential as a target for cancer intervention, we specifically knocked down AKR1B10 expression in colorectal cancer cells (HCT-8) and evaluated the resulting effect on cell proliferation. The results showed that downregulation of AKR1B10 evoked significant cell growth inhibition, DNA synthesis suppression and reduction of clonogenic growth capability. In contrast to AKR1B10 introduction,14 the silencing of this gene enhanced the cell susceptibility to reactive carbonyls, acrolein and crotonaldehyde. We investigated AKR1B10 silencing using two different siRNAs that target the encoding and 3′ untranslational regions of AKR1B10 gene, respectively. Both siRNAs exhibited identical experimental results, demonstrating the specificity of the AKR1B10 silencing results.

Clonogenic growth reflects the proliferative properties of an individual cell. The anchorage-independent growth in semisolid medium is a hallmark of cellular transformation/tumorigenicity and a basis of cancer cell invasion and metastasis.25, 26AKR1B10 silencing in HCT-8 cells reduced the growth rate by ∼50% and suppressed DNA synthesis up to nearly 40% compared with the control cells. Particularly, in anchorage-independent culture colony formation rates of AKR1B10 downregulated HCT-8 cells, defined as the percentage of formed clones over the number of plated cells, were decreased by ∼30%, and the sizes of the surviving colonies were dramatically diminished compared with the control cells. These data suggest that AKR1B10 may play a critical role in HCT-8 cancer cell proliferation, thus being a potential target for the intervention of cancer with AKR1B10 induction, such as HCC and lung SCC and adenocarcinomas.1, 12

Recombinant AKR1B10 protein showed high affinity and strong enzymatic activity toward acrolein and crotonaldehyde (Table I), indicating that AKR1B10 may participate in regulation of intracellular carbonyl levels. AR mediates cellular mitogenic signaling via controlling the concentration of intracellular carbonyl such as 4-hydroxynonenal.10 Whether AKR1B10 regulates cell proliferation by a similar mechanism needs to be further investigated. In addition, AKR1B10 is a dominant reductase of retinals, the precursors of retinoic acid signaling that regulates cell proliferation and differentiation.13 It is also possible that silencing of AKR1B10 gene altered this signaling pathway via accumulating retinals. Experiments are in progress to further elucidate the molecular mechanisms.

Exposure of HCT-8 cells with AKR1B10 knockdown to acrolein (25 μM) and crotonaldehyde (50 μM) resulted in rapid HCT-8 cell death (Figs. 3–5), indicating that AKR1B10 plays a critical role in regulating cellular response to additional carbonyl stress. In this study, acrolein and crotonaldehyde were used at greater concentrations to observe the acute toxicity because AKR1B10 silencing induced by chemically synthesized siRNAs is transitory (Fig. 1 in complementary data). Acrolein and crotonaldehyde are α, β-unsaturated aldehydes.18, 21, 27 These 2 compounds are strongly reactive due to α, β-unsaturated carbon–carbon bonds (via Michael addition) and carbonyl groups (via Schiff base), leading to growth inhibition, membrane permeability, glutathione (GSH) depletion and sulfhydryl oxidation.27, 28 Acrolein and crotonaldehyde are continuously produced by lipid peroxidation of polyunsaturated fatty acids.16, 17 Acrolein is also a product of myeloperoxidase-catalyzed amino acid oxidation17 and anticancer drug metabolism, such as cyclophosphamide.29 Living cells have developed several lines of defense mechanisms to eliminate these intracellular cytotoxins. Aldehyde dehydrogenases (ALDH) mediate the oxidation of aldehyde carbonyls, forming carbonic acids;30 glutathione-S-transferases (GST) catalyze the conjugation of carbonyls with glutathione;31, 32 and aldo–keto reductases such as AR participate in the reduction of carbonyl groups to alcohol forms.1, 33 Therefore, AR inhibitors have shown the capability of enhancing the sensitivity of cancer cells to reactive carbonyls and antitumor agents, becoming potent novel anticancer agents.6-9 Specific AKR1B10 inhibitors are not yet available, but identification of AKR1B10's role in HCT-8 cancer cell proliferation indicates the significance of developing AKR1B10 inhibitors as novel antitumor agents.

In normal tissues, AKR1B10 is specifically expressed in the small intestine and colon, 2 organs directly exposed to lumenal carbonyl toxins ingested from foodstuffs or locally produced by microbes.15, 34 AKR1B10's detoxification to carbonyl toxins implies its potential in protecting gastrointestinal tract from carbonyl toxins. Further study is warranted to elucidate the role of AKR1B10 in gastrointestinal tumorigenesis induced by reactive carbonyls.15, 20, 22

Finally, acrolein-induced HCT-8 cell death occurred with LDH efflux and annexin-V staining. Of the 3 forms of cell death (apoptosis, necrosis and oncosis), oncosis is featured with membrane disintegration and annexin-V staining via penetrating disintegrated cell membranes and binding to phospholipid phosphatidylserine on the inside.35 Therefore, HCT-8 cells with AKR1B10 silencing may experience an oncotic cell death process at greater concentrations of acrolein.14