Figures
Abstract
Although several Cre-loxP-based gene knockout mouse models have been generated for the study of gene function in alveolar epithelia in the lung, their applications are still limited. In this study, we developed a SPC-Cre-ERT2 mouse model, in which a tamoxifen-inducible Cre recombinase (Cre-ERT2) is under the control of the human surfactant protein C (SPC) promoter. The specificity and efficiency of Cre-ERT2 activity was first evaluated by crossing SPC-Cre-ERT2 mouse with ROSA26R mouse, a β-galactosidase reporter strain. We found that Cre-ERT2 was expressed in 30.7% type II alveolar epithelial cells of SPC-Cre-ERT2/ROSA26R mouse lung tissues in the presence of tamoxifen. We then tested the tamoxifen-inducible recombinase activity of Cre-ERT2 in a mouse strain bearing TSC1 conditional knockout alleles (TSC1fx/fx). TSC1 deletion was detected in the lungs of tamoxifen treated SPC-Cre-ERT2/TSC1fx/fx mice. Therefore this SPC-Cre-ERT2 mouse model may be a valuable tool to investigate functions of genes in lung development, physiology and disease.
Citation: Gui Y-S, Wang L, Tian X, Feng R, Ma A, Cai B, et al. (2012) SPC-Cre-ERT2 Transgenic Mouse for Temporal Gene Deletion in Alveolar Epithelial Cells. PLoS ONE 7(9): e46076. https://doi.org/10.1371/journal.pone.0046076
Editor: Zhiyuan Gong, National University of Singapore, Singapore
Received: April 15, 2012; Accepted: August 27, 2012; Published: September 25, 2012
Copyright: © Gui et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The work was supported by the 973 Program of China (2009CB522106). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Gene knockout technology has been widely used in mice to determine the functions of a specific gene in mouse development and its contribution to a particular disease. Conventional gene knockout ablates a gene in all cells, usually resulting in complex phenotypes, or embryonic lethality if the gene product is critical in development. Therefore, conditional or tissue-specific gene knockout strategies have been developed. One of these strategies is using a Cre-loxP recombination system. This system consists of a 38 kD Cre DNA recombinase and two 34 bp loxP sites flanking a target DNA sequence. The Cre recombinase recognizes loxP sites and excises the target DNA sequence when the orientation of two loxP sites is cis-repeated [1]. For a tissue-specific gene knockout strategy, the Cre coding sequence is usually driven by a tissue-specific promoter, which allows Cre to be expressed only in one type of tissue. Thus, deletion of the gene of interest also occurs only in the tissue where Cre is expressed.
Using Cre/loxP system to specifically knock out genes in alveolar epithelial cells has been reported by several research groups [2], [3]. In these reported Cre transgenic mouse models, a surfactant protein C (SPC) promoter is used to drive Cre expression. The surfactant protein C is exclusively expressed in the type II alveolar epithelial cells [4]. A SPC-Cre transgenic mouse was generated to understand the role of VEGF in alveolar structure, acute inflammation, and vascular permeability [2]. In this model, the Cre-mediated recombination begins at embryo period and is not controlled. This model is useful to study lung development at embryonic stage, but has limitations for study at adult stage.
To prevent the possible embryonic lethality caused by gene deletion in early embryo development, a SPC-rtTA/TetO-Cre transgenic mouse model has been developed [3]. In this model, Cre-recombinase activity can be controlled by doxycycline temporally. However, the breeding scheme for getting three transgenes in one mouse is very complicated and time-consuming [3]. In addition, the rtTA toxicity in SPC-rtTA mice leads to impaired alveologenesis, abnormal expression of surfactant-associated proteins, and mouse death, which limits the use of this model in research [5], [6].
An inducible Cre-recombinase, Cre-ERT2, has been developed by fusing the Cre coding sequence with a mutant form of ligand-binding domain of the estrogen receptor (ERT2). The Cre-ERT2 is active only in the presence of tamoxifen. A number of tissue-specific Cre-ERT2 mouse models have been reported [7], [8], [9]. In this study, we developed a novel SPC-Cre-ERT2 mouse model which can be applied for gene knock-out in mouse alveolar epithelium in a temporally controlled fashion. We tested this model in both ROSA26R and TSC1fx/fx transgenic mice.
Results
Generation of the SPC-Cre-ERT2 Transgenic Mice
A 9.5 kb AatII-NsiI DNA fragment containing SPC-Cre-ERT2 expression cassette (Fig. 1A) was microinjected into embryos of C57BL/6J mice. These embryos were transferred into pseudo-pregnant surrogate C57BL/6J mice to obtain pups. We used PCR to screen for mice bearing SPC-Cre-ERT2 and found five mice positive for Cre-ERT2 (Fig. 1B). These five founders were bred with C57BL/6J mice to transmit SPC-Cre-ERT2 transgene.
A) Schematic map of SPC-Cre-ERT2 expression cassette. HSPC-P, human surfactant protein C promoter; Cre-ERT2, Cre coding sequence fused with a tamoxifen-inducible estrogen receptor. pA, a polyA sequence from SV40 virus. Cre-F primer binding sites, 561–579 bp of Cre-ERT2 transgene; Cre-R primer binding sites, 976–997 bp of Cre-ERT2 transgene. The map is drawn in scale. B) Screening SPC-Cre-ERT2 transgenic mice using PCR. Genomic DNA from each mouse tail was used as template to specifically PCR-amplify the Cre-ERT2 transgene. M, DNA marker; +, SPC-Cre-ERT2 plasmid DNA control; -, water control; F8, F13, F16, F42, F67 are five representative founder transgenic mice generated by microinjection of SPC-Cre-ERT2 expression cassette into fertilized embryos. C) Cre-ERT2 fusion proteins were detected using Western blot in lung tissues of SPC-Cre-ERT2 transgenic mice. Lane 1, C57BL/6J mouse; Lane 2, an offspring of F42 founder without Cre-ERT2 transgene when genotyped using PCR; Lane 3, 4, 5, offspring of F42 founder; Lane 6, 7, 8, offspring of F67 founder. Notice the variable expression levels of Cre-ERT2 in offspring from the same founder.
Cre-ERT2 Expression in Lung Tissues
We then asked whether Cre-ERT2 fusion proteins were stably expressed in the lung tissues of the offspring of these five founders. Cre-ERT2 fusion proteins in the lung tissues of 4-week-old mice were analyzed using Western blot with monoclonal antibody against Cre. We found that Cre-ERT2 was expressed stably only in the offspring from two of these five founders (Fig. 1C). Therefore, the offspring from these two founders (F42 and F67) were used for the following experiments.
Cre-ERT2 Recombinase Activity in SPC-Cre-ERT2/ROSA26R Transgenic Mice
To test fusion protein, we asked whether Cre-ERT2 fusion protein could turn on LacZ gene expression in ROSA26R mice [10] in the presence of tamoxifen. The SPC-Cre-ERT2 mice were bred with ROSA26R mice to get SPC-Cre-ERT2/ROSA26R mice. The β-galactosidase activity was detected only in the alveolar epithelial cells of those mice receiving tamoxifen, but not vehicle, suggesting that the Cre-ERT2 fusion protein was functional in the alveolar epithelial cells of mice in the presence of tamoxifen (Fig. 2A).
A) β-galactosidase activity was detected in the lung alveolar epithelia from a SPC-Cre-ERT2/ROSA26R transgenic mouse receiving tamoxifen treatment. a & b, lung tissue sections from ROSA26R mice (without Cre-ERT2 transgene) receiving vehicle (a) or tamoxifen (b); c & d, lung tissue sections from SPC-Cre-ERT2/ROSA26R transgenic mice receiving vehicle (c) or tamoxifen (d). B) Endogenous β-galactosidase activity was found in lung bronchial epithelia of ROSA26R mouse receiving vehicle (a) and SPC-Cre-ERT2/ROSA26R transgenic mouse after tamoxifen treatment (b). C) X-gal stained lung tissues of SPC-Cre-ERT2/ROSA26R transgenic mouse receiving tamoxifen were also immune-stained for proSP-C, an alveolar type II cell-specific marker. Arrows indicate three representative alveolar type II cells co-expressing proSP-C (brown) and β-galactosidase (blue). D) β-galactosidase activity was detected only in the lung alveolar epithelium, but not in other organs from a SPC-Cre-ERT2/ROSA26R transgenic mouse receiving tamoxifen treatment. a, heart; b, liver; c, kidney; d, intestine; e, lung. All scale bars in this figure equal 100 µm.
We then asked the tamoxifen-induction efficiency of the Cre-ERT2 in type II alveolar epithelial cells by using immunohistochemical staining against alveolar type II cell-specific marker protein, proSP-C, and X-gal staining in the lung tissues of 4-week-old SPC-Cre-ERT2/ROSA26R mice receiving tamoxifen (1 mg/day) for five consecutive days. About 30.7% alveolar type II epithelial cells were β-galactosidase positive (Fig. 2C).
We also investigated the specificity of the SPC-Cre-ERT2 in the lung tissues of those SPC-Cre-ERT2/ROSA26R mice. Endogenous β-galactosidase staining was found in bronchial epithelia from ROSA26R mice receiving vehicle and SPC-Cre-ERT2/ROSA26R mice with tamoxifen treatment (Fig. 2B). We found that, in lung tissue, excluding bronchial epithelial cells, 51.8% of β-galactosidase positive cells were alveolar type II cells (Fig. 2C). Furthermore, the tamoxifen-induced recombinase activity of Cre-ERT2 existed only in lung, but not in heart, liver, kidney, and intestine, suggesting that the SPC promoter conveyed tissue-specific activity in the SPC-Cre-ERT2 transgene (Fig. 2D).
Cre-ERT2 Recombinase Activity in SPC-Cre-ERT2/TSC1fx/fx Transgenic Mice
We also tested the tamoxifen-inducible recombinase activity of Cre-ERT2 in the TSC1fx/fx transgenic mice [11]. The SPC-Cre-ERT2/TSC1fx/fx mice were obtained by breeding SPC-Cre-ERT2 with TSC1fx/fx mice as described in the section of materials and methods. After given tamoxifen or vehicle, the SPC-Cre-ERT2/TSC1fx/fx mice were sacrificed, their lung tissues were harvested, and their genomic DNAs were extracted to detect TSC1 deletion by PCR [11]. We found that TSC1 deletion was detected only in mice treated with tamoxifen, but not vehicle, suggesting that the Cre-ERT2 exerted its recombinase activity only in the presence of tamoxifen (Fig. 3).
DNAs from the lung tissues of SPC-Cre-ERT2/TSC1fx/fx transgenic mice treated with vehicle or tamoxifen were examined by PCR to detect TSC1 deletion as an indication of Cre-ERT2 recombinase activity. M, DNA marker; +, a transgenic mouse with TSC1 deletion; -, C57BL/6J mouse. Other lanes, offspring from F42 and F67 founders treated with vehicle (vehi) or tamoxifen (tam).
Discussion
In this study, we generated a SPC-Cre-ERT2 transgenic mouse model which can be used to study the functions of genes in mouse alveolar epithelial cells in any stage of mouse development. We tested this model in two commonly used transgenic mouse strains carrying either ROSA26R [10] or TSC1 conditional knockout alleles (TSC1fx/fx) [11] and found that this novel transgenic mouse model expressed Cre-ERT2 fusion proteins predominately in alveolar epithelial cells and Cre recombinase activity was active only in the presence of tamoxifen (Fig. 2 and 3).
Our SPC-Cre-ERT2 transgenic mouse model has several advantages when compared with other transgenic mouse models which have been used for conditionally knocking out genes for lung development and functions [2], [3]. First, in our transgenic mouse model we can control when Cre recombination takes place by tamoxifen administration, which allows scientists to knock out the gene of interest in a temporally controlled manner. Second, using single Cre-ERT2 transgene in our transgenic mouse model makes the subsequent mouse breeding process simpler than using two transgenes in the SPC-rtTA/TetO-Cre mouse model [3].
We observed variable Cre-ERT2 expression in our SPC-Cre-ERT2 transgenic mice. Three out of five SPC-Cre-ERT2 founders could not give birth of mice expressing Cre-ERT2 in lungs. Furthermore, the Cre-ERT2 expression levels in the lungs of offspring from the same founders were variable (Fig. 1C). We also observed the different degree of β-galactosidase staining in the lungs of mice from the same founder line. One possible reason for the difference of transgene expression levels in offspring from the same founder line might be that multiple integration sites occurred in the founder mice.
In a healthy human lung, 15% cells are alveolar type II epithelial cells [12]. In our study, an average of 30.7% alveolar type II cells in the lungs of the SPC-Cre-ERT2/ROSA26R mice treated with tamoxifen expressed β-galactosidase. This may be due to the lack of an optimal condition for inducing Cre-ERT2 by tamoxifen in our transgenic mice. The optimal conditions for inducing gene inactivation by administrating tamoxifen are different in different mouse models [7], [8], [9], [13], [14]. An optimal condition per mouse model can be affected by the following factors, the means of drug administration, the dose of tamoxifen, and the age, gender, and genetic background of the experimental mice. Further studies to determine an optimal tamoxifen-induction protocol for our transgenic mouse model are necessary for its wide application.
Another discrepancy in our study was that only 51.8% β-galactosidase positive cells (excluding the β-galactosidase positive bronchial epithelia) in lung tissues of SPC-Cre-ERT2/ROSA26R mice receiving tamoxifen were alveolar type II cells (Fig. 2C). Other β-galactosidase positive cells were bronchial epithelial cells (Fig. 2B–b) and a few of other types of cells. The β-galactosidase positive cells in bronchial epithelium were also observed in ROSA26R mice (Fig. 2B-a) or SPC-Cre-ERT2/ROSA26R mice without tamoxifen treatment (data not shown). Thus, the β-galactosidase activity in bronchial epithelium is caused by its endogenous activity. A few cells which were β-galactosidase positive and proSP-C negative were observed in the lung tissues of our experimental mice. We did not further clarify these cells. Some of them might be alveolar type I cells because of the alveolar type II cells to type I cells transdifferentiation [15].
In summary, we generated a novel transgenic mouse model, SPC-Cre-ERT2 mice, which express tamoxifen-inducible Cre recombinase in alveolar epithelial cells. This mouse model can be used to conditionally knock out genes in alveolar epithelial cells at any stages of lung development. Therefore, it will be a useful tool in understanding gene functions in lung development, physiology and disease.
Materials and Methods
Ethics Statement
All animal experiments were approved by the Animal Ethics Committee of Peking Union Medical College according to international and institutional guidelines for animal care. All surgery was performed under sodium pentobarbital anesthesia, and all efforts were made to minimize suffering.
Plasmid and Transgene Construction
A SPC-Cre-ERT2 expression construct (Fig. 1A) was generated as following steps. First, a 3.8-kb AatII/SpeI human surfactant protein C promoter (HSPC-P) from pUC-SPC plasmid (a gift from Dr. Jeffrey A. Whitsett, Cincinnati Children’s Hospital Medical Center) was cloned to a pGEM-TEasy vector (Promega, USA) to generate pGEM-SPC. Then, a 2.8-kb NotI/NsiI fragment containing Cre-ERT2 cDNA and the polyadenylation site (poly A) from the SV40 early region from a pCAG-Cre-ERT2 (Addgene, USA) [16] was cloned to the pGEM-SPC to get pGEM-SPC-Cre-ERT2 (Fig. 1A).
Generation of Transgenic Mice
To generate SPC-Cre-ERT2 transgenic mice, a 9.5-kb AatII/NsiI fragment from pGEM-SPC-Cre-ERT2 construct was microinjected into fertilized embryos of C57BL/6J mice. The SPC-Cre-ERT2 founder lines were bred with C57BL/6J wild type mice to generate the SPC-Cre-ERT2 F1 mice. Offspring were genotyped as described below. Mice were maintained in a pathogen-free mouse facility.
Western Blot Analysis for Cre-ERT2 Expression
Total proteins were extracted from mouse lung tissues and separated on a 12% SDS-PAGE gel. The separated proteins were then transferred onto a PVDF membranes (Pierce, USA) for detecting Cre-ERT2 and β-actin proteins with monoclonal antibodies against Cre (1∶1,000, Covance, USA) and β-actin (1∶1,000, Santa Cruz, USA), respectively. The β-actin expression was used as a loading control of protein extracts.
Analysis of Cre-ERT2 Activity
We analyzed Cre-ERT2 recombinase activity in two transgenic mouse models: ROSA26R and TSC1fx/fx mice. ROSA26R mouse strain bear a lox-STOP-lox-cassette inserted between a promoter and a LacZ gene [10] and TSC1fx/fx mice bear loxP-flanked TSC1 conditional knockout alleles [11]. The SPC-Cre-ERT2/ROSA26R mice were obtained via breeding SPC-Cre-ERT2 mice with homozygous ROSA26R mice. The SPC-Cre-ERT2/TSC1fx/fx mice were obtained via two steps of breeding. First, the SPC-Cre-ERT2 transgenic mice were bred with TSC1fx/fx mice to get SPC-Cre-ERT2/TSC1fx/+ mice. Then these SPC-Cre-ERT2/TSC1fx/+ mice were bred with TSC1fx/fx mice to get SPC-Cre-ERT2/TSC1fx/fx mice. All transgenic mice were genotyped using PCR. Primers used for amplifying Cre transgene were forward primer, 5′-TACTGACGGTGGGAGAATG-3′ and reverse primer, 5′-CTGTTTCACTATCCAGGTTACG-3′. The expected PCR product was 437 bp in size. Primers used for amplifying ROSA26R were forward primer, 5′-GACACCAGACCAACTGGTAATGGTAGCGAC-3′ and reverse primer, 5′-GCATTGAGCTGGGTAATAAGCGTTGGCAAT-3′. The expected PCR product was 750 bp in size. The TSC1fx/fx conditional allele was tested using PCR as previously described [11].
Cre-ERT2 recombinase activity in these transgenic mice was induced by tamoxifen [9], [14]. Briefly, tamoxifen (Sigma, USA) was dissolved in a sunflower seed oil/ethanol mixture (10∶1) at 10 mg/ml. Each 4-week-old mouse was injected intraperitoneally with 100 µl of tamoxifen per day for 5 consecutive days. One week after the last injection, mice were sacrificed and the lungs, hearts, livers, kidneys, intestines of the SPC-Cre-ERT2/ROSA26R mice were harvested to examine the β-galactosidase activity by X-gal staining and the lungs of SPC-Cre-ERT2/TSC1fx/fx mice were harvested to detect the TSC1 deletion.
For β-galactosidase activity analysis, tissues were fixed in 0.2% glutaraldehyde, 2% formaldehyde in PBS for overnight at 4°C. After washed three times for 45 min with PBS at room temperature with gentle shaking, tissue sections were stained overnight in freshly prepared X-gal staining solution containing 2 mM MgCl2, 2.5 mM K3Fe(CN)6, 2.5 mM K4Fe(CN)6 and 1 mg/ml X-gal. X-gal stock-solution was prepared in dimethylformamide at 40 mg/ml. Stained tissues were further paraffin embedded and sections were counterstained with nuclear fast red.
To investigate the Cre-ERT2-mediated excision efficiency in the lung tissues, we performed immunohistochemical staining for pro-Surfactant Protein C (proSP-C), a specific alveolar type II cell marker and β-galactosidase staining at the same lung tissues. Briefly, X-gal stained lung tissue sections (3 µm) were deparaffinized and rehydrated in xylene, heated in 10 mM citrate buffer, treated with 3% H2O2 for 10 minutes. After blocked with 5% normal goat serum, these tissue sections were incubated with rabbit-derived primary antibody (anti-mouse proSP-C, 1∶1000, Millipore, USA) at 4°C for overnight followed by incubation with secondary antibody for 1 hour at room temperature. The bound secondary antibodies were detected with VECTASTAIN ABC kit (Vector, USA). Images were taken with Nikon Eclipse 80i microscope (Tokyo, Japan) with a digital camera. The numbers of cells positive for both β-galactosidase and proSP-C, β-galactosidase only, or proSP-C only were obtained respectively by counting total 12 fields with 400×magnification.
For detection of TSC1 deletion, genomic DNAs were isolated from lung tissues of SPC-Cre-ERT2/TSC1fx/fx mice treated with tamoxifen. The TSC1 deletion was detected using PCR. The primers for TSC1 deletion were: F4536∶5′-AGGAGGCCTCTTCTGCTACC-3′ and R6548∶5′-TGGGTCCTGACCTATCTCCTA-3′ [11]. The expected PCR product size was 370 bp with excision of exons 17 and 18 of TSC1 gene [11].
Acknowledgments
We thank Dr Jeffrey A. Whitsett at the Cincinnati Children’s Hospital Medical Center, the University of Cincinnati College of Medicine for his kindly providing the pUC-SPC plasmid. We also thank the Institute of Laboratory Animal Science of Chinese Academy of Medical Sciences for carrying out the plasmid microinjection.
Author Contributions
Conceived and designed the experiments: KFX HZ. Performed the experiments: YSG LW. Analyzed the data: YGS RF KFX. Contributed reagents/materials/analysis tools: XT AM BC. Wrote the paper: YSG KFX.
References
- 1. Nagy A (2000) Cre recombinase: the universal reagent for genome tailoring. Genesis 26: 99–109.
- 2. Mura M, Binnie M, Han B, Li C, Andrade CF, et al. (2010) Functions of type II pneumocyte-derived vascular endothelial growth factor in alveolar structure, acute inflammation, and vascular permeability. Am J Pathol 176: 1725–1734.
- 3. Sun J, Chen H, Chen C, Whitsett JA, Mishina Y, et al. (2008) Prenatal lung epithelial cell-specific abrogation of Alk3-bone morphogenetic protein signaling causes neonatal respiratory distress by disrupting distal airway formation. Am J Pathol 172: 571–582.
- 4. Tafel O, Latzin P, Paul K, Winter T, Woischnik M, et al. (2008) Surfactant proteins SP-B and SP-C and their precursors in bronchoalveolar lavages from children with acute and chronic inflammatory airway disease. BMC Pulm Med 8: 6.
- 5. Morimoto M, Kopan R (2009) rtTA toxicity limits the usefulness of the SP-C-rtTA transgenic mouse. Dev Biol 325: 171–178.
- 6. Whitsett JA, Perl AK (2006) Conditional control of gene expression in the respiratory epithelium: A cautionary note. Am J Respir Cell Mol Biol 34: 519–520.
- 7. Monvoisin A, Alva JA, Hofmann JJ, Zovein AC, Lane TF, et al. (2006) VE-cadherin-CreERT2 transgenic mouse: a model for inducible recombination in the endothelium. Dev Dyn 235: 3413–3422.
- 8. Lantinga-van LI, Leonhard WN, van de Wal A, Breuning MH, Verbeek S, et al. (2006) Transgenic mice expressing tamoxifen-inducible Cre for somatic gene modification in renal epithelial cells. Genesis 44: 225–232.
- 9. El MF, Janssen KP, Chang BH, Li M, Hindie V, et al. (2004) Tissue-specific and inducible Cre-mediated recombination in the gut epithelium. Genesis 39: 186–193.
- 10. Mao X, Fujiwara Y, Orkin SH (1999) Improved reporter strain for monitoring Cre recombinase-mediated DNA excisions in mice. Proc Natl Acad Sci U S A 96: 5037–5042.
- 11. Kwiatkowski DJ, Zhang H, Bandura JL, Heiberger KM, Glogauer M, et al. (2002) A mouse model of TSC1 reveals sex-dependent lethality from liver hemangiomas, and up-regulation of p70S6 kinase activity in Tsc1 null cells. Hum Mol Genet 11: 525–534.
- 12. Crapo JD, Barry BE, Gehr P, Bachofen M, Weibel ER (1982) Cell number and cell characteristics of the normal human lung. Am Rev Respir Dis 126: 332–337.
- 13. Santagati F, Minoux M, Ren SY, Rijli FM (2005) Temporal requirement of Hoxa2 in cranial neural crest skeletal morphogenesis. Development 132: 4927–4936.
- 14. Kuhbandner S, Brummer S, Metzger D, Chambon P, Hofmann F, et al. (2000) Temporally controlled somatic mutagenesis in smooth muscle. Genesis 28: 15–22.
- 15. Fehrenbach H (2001) Alveolar epithelial type II cell: defender of the alveolus revisited. Respir Res 2: 33–46.
- 16. Indra AK, Warot X, Brocard J, Bornert JM, Xiao JH, et al. (1999) Temporally-controlled site-specific mutagenesis in the basal layer of the epidermis: comparison of the recombinase activity of the tamoxifen-inducible Cre-ER(T) and Cre-ER(T2) recombinases. Nucleic Acids Res 27: 4324–4327.