http://orcid.org/0000-0001-8785-5439
Figures
Abstract
There is an increasing demand for genetically modified mice produced without crossing, for rapid phenotypic screening studies at the organismal level. For this purpose, generation of completely embryonic-stem-cell (ESC)-derived chimeric mice without crossing is now possible using a microinjection or aggregation method with 3i culture medium. However, the microinjection of ESCs into blastocyst, morula, or 8-cell-stage embryos requires a highly skilled operator. The aggregation method is an easier alternative, but the conventional aggregation protocol still requires special skills. To make the aggregation method easier and more precise, here we developed a micro-aggregation device. Unlike conventional 3-dimensional culture, which uses hanging-drop devices for aggregation, we fabricated a polystyrene funnel-like structure to smoothly drop ESCs into a small area (300-μm in diameter) at the bottom of the device. The bottom area was designed so that the surface tension of the liquid-air interface prevents the cells from falling. After aggregation, the cells can be recovered by simply exerting pressure on the liquid from the top. The microdevice can be set upon a regular 96-well plate, so it is compatible with multichannel pipette use or machine operation. Using the microdevice, we successfully obtained chimeric blastocysts, which when transplanted resulted in completely ESC-derived chimeric mice with high efficiency. By changing the number of ESCs in the aggregate, we found that the optimum number of co-cultured ESCs was around 90~120 per embryo. Under this condition, the efficiency of generating completely ESC-derived mice was the same or better than that of the injection method. These results indicated that our microdevice can be used to produce completely ESC-derived chimeric mice easily and with a high success rate, and thus represents a promising alternative to the conventional microinjection or aggregation method, especially for high-throughput, parallel experimental applications.
Citation: Sumiyama K, Matsumoto N, Garçon-Yoshida J, Ukai H, Ueda HR, Tanaka Y (2018) Easy and efficient production of completely embryonic-stem-cell-derived mice using a micro-aggregation device. PLoS ONE 13(9): e0203056. https://doi.org/10.1371/journal.pone.0203056
Editor: Qiang Wu, Macau University of Science and Technology, MACAO
Received: March 22, 2018; Accepted: August 14, 2018; Published: September 19, 2018
Copyright: © 2018 Sumiyama 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.
Data Availability: All relevant data are within the paper.
Funding: The study was supported by JSPS KAKENHI Grant Numbers JP15H05949 (KS), https://kaken.nii.ac.jp/en/grant/KAKENHI-PLANNED-15H05949/ and JP15H04408 (KS), https://kaken.nii.ac.jp/en/grant/KAKENHI-PROJECT-15H04408/, and the strategic programs for R&D (President’s discretionary fund) of RIKEN (YT, HRU) http://www.riken.jp. 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
Mouse genome editing is an indispensable process for organism-level phenotype analysis in biology and medical research. The CRISPR/Cas9 system [1, 2] makes it possible to directly edit the genome of mouse fertilized eggs, generating mutations by homologous recombination (HR) without using embryonic stem cells (ESCs) [3, 4]. The method for directly injecting the CRISPR/Cas9 components into fertilized eggs is easy and efficient for introducing small mutations. Although the use of ESCs is less efficient than the direct introduction of mutations using Cas9 [4, 5], it still has some advantages, such as the ability to generate knock-ins of a relatively large size. Furthermore, knock-in generation by the direct injection method often results in mosaic individuals, which can be problematic for performing phenotype analysis at the F0 generation. Mutant ESCs can be easily maintained and stored without generating mice. Thus, it is possible to retain mutations that cause lethality in later development, and further rescue experiments can be performed in an ESC strain with a lethal mutation [6, 7]. Because of these advantages, ESCs are still used to produce genetically modified mice.
There are several methods for generating individual mice from established ESC lines with targeted mutations. The direct injection of ESCs into blastocysts is commonly used [8]. However, when C57BL/6-lineage ESCs (black coat color) are injected into host embryos such as ICR (white coat color), chimeras can be obtained with different degrees of mosaic coat color. Thus, further crossbreeding is usually necessary to obtain a completely ESC-derived mutant line. Since this process takes nearly a year, it is often the rate-determining step of experiments using knockout or knock-in mice. Recently, improved control of the ESC differentiation state by adding special inhibitors to the culture medium (3i culture method) for ESC establishment and subsequent culture passages, combined with the injection of ESCs into 8-cell-stage embryos rather than blastocysts, enabled the direct generation of completely ESC-derived chimeric mice (Fig 1A) [6, 7]. This procedure makes it possible to analyze mouse phenotypes without crossing, which is highly beneficial for the high-throughput screening of mutation phenotypes on a large scale.
(A) Injection method: completely separated ESCs are microinjected into 8-cell-stage embryos. (B) Conventional aggregation method: a clump of ESCs subjected to partial trypsinization are co-cultured with 8-cell-stage embryos with the zona-pellucida removed, in a dent made in the bottom of a culture dish. (C) Microaggregation method: ESCs and 8-cell-stage embryos, both prepared as in B, are separately introduced from the top of the device into a micro-hanging drop and co-cultured. In all three methods, chimeric embryos are cultured until the blastocyst stage and then transplanted into the uterus of recipient mice.
The injection method can produce very good results when a highly skilled operator performs it using a dedicated injection device [6, 7]. However, because it can be difficult to acquire this injection skill and the apparatus is expensive, the injection method cannot be performed in any laboratory. Therefore, as an alternative, a chimeric embryo preparation method using aggregation was established for C57BL/6N ESCs [8–10]. In this procedure, a small indentation is made with a needle on the bottom of a plastic culture dish, and the embryo is brought into contact with a clump of ESCs in the dent (Fig 1B). After incubating this co-culture overnight, embryos containing ESCs are generated. This method produces chimeric mice as efficiently as the injection method. However, it still requires special skills to prepare appropriately shaped dents on the culture dish bottom and to carefully place and recover the embryos and ESCs, to obtain good and reproducible results. It is also difficult to control the number of ESCs that aggregate with the embryo, so it is challenging to determine the optimal quantity of ESCs for the best results. To solve these problems, here we sought to develop a dedicated microdevice for aggregation that could be easily and widely applied to produce completely ESC-derived mice with high reproducibility (Fig 1C).
From the viewpoint of device engineering, a number of reports on cell culture and handling in microfluidic devices have appeared in the past decade or so [11–13]. Most of them are 2-dimensional (2D) culture devices for growing cells on a flat substrate for high-throughput analysis [14]. However, these conditions do not resemble the 3-dimensional (3D) in vivo environment [15]. “Body-on-a-chip” or “organ-on-a-chip” 3D-culture systems are recent trends in the “lab-on-a-chip” field [15–18]. To obtain a 3D culture, one major approach is to form spheroids by the air-liquid curved interface in a hanging-drop [19, 20]. Although this method is simple and the recovery is relatively easy, the droplet diameter is more than 1 mm, which results in an air-liquid interface with a relatively large curvature that may not be suitable for the very small number of cells used in ESC–embryo aggregates. To fabricate these large curvature structures, semi-spherical small micro-chamber arrays on a chip have been reported [21, 22]. Although these systems are useful for high-throughput screening, it is difficult to recover the cells after spheroid formation. A system using droplets containing small numbers of cells was also reported [23]; however, in this case the droplet diameter is too small to accommodate large cell masses such as embryos. Thus, in summary, microdevices that can both aggregate cells and recover them well have not yet been realized. To overcome these problems, here we developed a new, very simple micro hanging drop system for mouse ESC-embryo aggregation that is made of polystyrene in 96-well culture-plate format. The fundamental concept of the device compared to conventional methods is summarized in Fig 1.
Materials and methods
Animal experiments
ICR mice were purchased from Oriental Yeast Co., Japan. All mice were given food and water ad libitum. Animals were kept in an SPF facility with a 12-hour light and 12-hour dark cycle (lights on at 8:00 am). The ambient temperature was kept around 21 degrees Celsius with a relative humidity of 50%. ICR mice (12 to 20 weeks) were used as recipients. A combination anesthetic (0.75 mg/kg of medetomidine, 4.0 mg/kg of midazolam, and 5.0 mg/kg of butorphanol) was used for surgery. The anesthetics were administered to recipient mice by intraperitoneal injection. All animal experiments were approved by the Institutional Animal Care and Use Committee of the RIKEN Kobe branch (approval number: QA2013-04-4).
ES cell culture
The ES cell line (R26-H2B-EGFP/mCherry KIES) was established from an F1 mouse generated by crossing R26-H2B-EGFP (CDB0238K) and R26-H2B-mCherry (CDB0239K) provided by LARGE, RIKEN CLST [24]. ES cells were established as described previously [7]. Feeder-free ES cell culture was performed as described previously [25]. Three days after starting the ESC culture, a single-cell suspension of ESCs was prepared as described previously [6]. After centrifugation, the ESC pellet was washed with PBS and re-suspended in a small amount of KSOM medium by gently pipetting several times. The concentration was adjusted with KSOM to 3 x 104 cells per ml, then used for aggregation experiments.
Mouse embryo preparation and transplantation
The procedures for obtaining, culturing, and transplanting embryos were performed according to standard protocols [8]. Briefly, host ICR embryos were recovered from ICR females at 2.5 days post-coitum and further cultured in KSOM medium until they reached the 8-cell stage. Embryos were treated in acidified Tyrode’s solution (pH 2.5, SIGMA T1788) until the zona pellucida was completely removed, then quickly moved to M2 medium. After aggregation was completed, blastocyst embryos were recovered and transferred into the uterus of pseudopregnant ICR females.
Device preparation
The device was made by injection molding, a manufacturing process in which parts are produced by injecting material into a mold, following a standard protocol [26], and this procedure was carried out in a manufacturing company (Fujimori Sangyo Co., Ltd, Japan) based on the device design prepared using a 3-dimensional design software (CADmeister, Nihon Unisys, Japan). Briefly, a stainless steel mold was fabricated by precision machining to form the features of the desired part. Then, melted polystyrene at over 170°C was poured into the mold. After cooling to solidify the polystyrene, the final product was removed from the mold. The polystyrene device was then sterilized by a γ beam at 10 kGy. After sterilization, the products were manually set onto each well of a commercially available 96-well polystyrene plate (Nunc) and packed until used for cell culture experiments.
Results and discussion
Device design and principle
The microdevice is compatible with a general-purpose 96-well plate, and it can easily be used with multichannel pipetting instruments and machines such as automatic pipetting devices and cell sorters. The device is designed to fit into the wells of a round-bottom 96-well plate. The device is funnel-shaped, with a wide, open upper part that narrows to a capillary at the bottom with an inner diameter of 300 μm (Fig 2A). When the device is filled with medium, aggregation can be performed by co-culturing the embryo and the ESCs at the opening end of the capillary's lower opening, in a minimally sized hanging drop. Because of the steep funnel shape, the introduced ESCs and embryos arrive at the lower part without sticking to the wall, and gather naturally along the curved surface of the liquid bottom (Fig 2A). The number of ESCs introduced can be easily controlled by adjusting the concentration of ESCs in the medium. To prevent evaporation, medium is placed in the well bottom, and the upper part of the device is overlaid with mineral oil (Fig 2A). The device is then placed in an incubator overnight. After incubation, the cells and embryo are collected by pushing the medium downward from the top, using a pipet tip, and collected in the bottom well. It is also possible to recover the aggregates by centrifuging the plate.
(A) Cross-sectional schematic view of the cell-aggregation device and the procedure for recovering the ESC-fused embryo using the device. (B) Image of designed devices set on a polystyrene 96-well plate. (C) Metal mold for polystyrene casting. (D) Picture of a fabricated device. (E) A polystyrene 96-well plate equipped with ESC aggregation devices. (F) An enlarged side-view picture of polystyrene devices installed on the 96-well plate.
The design of the devise on a plate is shown in Fig 2B, and the mold and fabricated devices are shown in Fig 2C–2F.
Parameter determination
The design of the device was based on the following analysis (also see Fig 3). To maintain the air-liquid interface without losing liquid into the reservoir, the surface tension-based force must be greater than the gravity-based force. The gravity force applied to the medium in the device (Fh) was calculated by the following equation:
(Eq 1)
where r is the radius of the nozzle of the device, ρ is the density of the liquid, g is the acceleration rate of gravity, and h is the distance from the bottom to the top of the liquid. On the other hand, the surface tension-based force on the medium is calculated by the following equation:
(Eq 2)
where ɣ is the surface tension of the liquid and ɵ is the contact angle between the liquid surface and the device material (see Fig 3). Eq (1) indicates that Fh is proportional to r2, while Eq (2) indicates that Fs is proportional to r. Therefore, as r decreases, Fh becomes less than Fs. In this calculation, the following physical property values were known: ρ = 1.0×103 kg/m3, g = 9.8 m/s2, and ɣ = 72 mN/m (around room temperature) [27]. The contact angle (ɵ) was determined by pictures of water droplets on the polystyrene material to be 81°±3° (n = 3, error indicates standard deviation). For easy liquid handling and to keep enough medium in the device so that the concentration does not change, the device height (h) was designed to be 7.0 mm. Using these values in Eqs (1) and (2), the nozzle radius (r) required to make Fh = Fs was calculated to be 0.30 mm. Because some turbulence and vibration would occur during the actual experiments, to ensure that the drop would not fall, r was designed to be half the calculated maximum value, namely 0.15 mm.
Results of aggregation experiments
To confirm that the device is suitable for embryonic culture, pronuclear-stage embryos were first placed in the device filled with KSOM medium and cultured for 2 days. The embryos were alive and developed to the 4-cell stage after incubation. This result indicated that embryos could be safely cultured in the device for at least 2 days (Fig 4).
Six pronuclear-stage embryos were cultured in the microdevice. After two days of culture, the embryos were alive and had developed normally. Embryos were observed from the top of the microdevice using a stereoscopic microscope.
Next, mouse production experiments using recombinant ESCs (R26-H2B-EGFP / mCherry KIES) were performed. The ESCs were established and maintained in 3i medium (iSTEM medium, Cellartis, Takara Bio, Japan). Eight-cell-stage embryos of ICR mice (Oriental Yeast Co., Japan) were used as the aggregation host embryos. After the device was filled with a small amount of KSOM, a single embryo with the zona pellucida removed was placed in each device, and then ESCs suspended in medium were placed in the device (Fig 2A). The optimal number of ESCs was examined using 60, 90, 150, and 300 cells per well. Twenty-four 8-cell-stage embryos were used for each ESC number condition. After culturing overnight, the embryos were collected from the device. The number of embryos recovered was 20 (60 ESC condition), 22 (90 ESC condition), 24 (150 ESC condition), and 23 (300 ESC condition). The recovered embryos were then transferred into the uterus of recipient ICR pseudopregnant female mice (SLC), and newborns were obtained by Caesarean section. The number of newborn pups obtained from each group was 2 (60 ESC condition), 7 (90 ESC condition), 5 (150 ESC condition), and 1 (300 ESC condition). Among them, 1 (60 ESC condition), 6 (90 ESC condition), 5 (150 ESC condition), and 1 (300 ESC condition) were judged to be completely ESC-derived mice from their coat color (Fig 5). Thus, under each condition we obtained a high proportion of completely ESC-derived mice. Note that not all but just a small fraction of the co-cultured ESCs contributed to the aggregated embryo, given that many ESCs remained outside of the blastocyst when the aggregation was completed. These results showed that the optimum number of ESCs co-cultured per well should be around 90 to 120 per embryo.
Pups labeled ES60, ES90, ES150, and ES300 were derived from embryos aggregated in the microdevice with 60, 90, 150, and 300 ESCs, respectively. ESV60 pups were derived from embryos aggregated with 60 ESCs in the bottom of a V-bottomed 96-well culture plate. All pups obtained are pictured except for one ES150 pup that had died.
To confirm that the ESCs used in the aggregation experiments would also work using a conventional method, we subjected 8-cell-stage embryos (Fig 1A) to injection with the same ESCs as used in the aggregation experiment. Ten ESCs were injected into each 8-cell-stage embryo, and a total of 87 embryos surviving after overnight culture were used for uterus transplantation. The resulting number of newborns was 21, of which 15 were completely ESC-derived mice (Table 1). In this conventional method, we obtained a 71% success rate of completely ESC-derived mice, whereas in the aggregation method using the newly developed microdevice, we achieved 50% to 100% (87% in total), which was comparable to or better than that achieved with the well-established injection method.
To test whether ESC-derived germline cells were transmitted to the next generation, several male mice obtained using the microdevice aggregation method that had a completely black coat were selected and mated with ICR female mice. Notably, all of the resulting newborns had a black coat, indicating successful germline transmission, with no contamination of ICR host cells in the germline (Fig 6, Table 2).
Left: F1 pups from ID ES90-#2. Right: F1 pups from ID ES300-#1.
Our results collectively indicated that the aggregation method using the microdevice can produce completely ESC-derived mice with an efficiency comparable to or even better than that of the 8-cell-stage ESC injection method, with a much simpler operation. With this device, researchers can easily produce completely ESC-derived mice without acquiring special operating skills. Furthermore, since a 96-well plate is used, experimental manipulation by a robot can be easily performed, and a high-throughput experimental system can be implemented. That is, combined with automation, this method is scalable and suitable for high-throughput operations.
We also compared this system with larger devices designed to produce conventional embryoid bodies. First, the hanging drop culture plate for preparing the embryoid body was excluded from consideration, because the size of the drop was much larger and the curvature of the liquid surface was too low to enable the ESCs and the embryo to aggregate well. Second, we examined a commercially available V-bottom 96-well culture plate. We tested and confirmed that embryoid bodies could be prepared in this plate, and then co-cultured 60 ESCs and an 8-cell-stage embryo in each well. Of 24 co-cultures performed, 21 embryos were obtained after overnight culture, and uterus transplantation was performed. From these embryos, four newborns were obtained, but all of them were chimeras with a low ESC contribution, as indicated by a white or mosaic coat color (Table 1, Fig 5). This result showed that the production of ESC-derived mice by aggregation using a conventional large-sized culturing apparatus with a low-curvature bottom surface is not very efficient. In contrast, our newly developed micro-aggregation device with a micro-hanging drop system is broadly applicable and highly effective for producing completely ESC-derived mice with ease.
Supporting information
S1 Checklist. The ARRIVE guidelines checklist.
Completed “The ARRIVE Guidelines Checklist” for reporting animal data in this manuscript.
https://doi.org/10.1371/journal.pone.0203056.s001
(PDF)
Acknowledgments
We thank Ms. A. Tsujii and Dr. H. Moriguchi at RIKEN, Japan for their fundamental investigation toward the design of the device. We thank M. Muramatsu for help in establishing ESCs, and S. Morino for help in maintaining the ESCs. We also thank Fujimori Sangyo Co., Ltd, Japan for their help in designing the device.
References
- 1. Cong L, Ran FA, Cox D, Lin S, Barretto R, Habib N, et al. Multiplex genome engineering using CRISPR/Cas systems. Science. 2013;339(6121):819–23. pmid:23287718
- 2. Mali P, Yang L, Esvelt KM, Aach J, Guell M, DiCarlo JE, et al. RNA-guided human genome engineering via Cas9. Science. 2013;339(6121):823–6. pmid:23287722
- 3. Wang H, Yang H, Shivalila CS, Dawlaty MM, Cheng AW, Zhang F, et al. One-step generation of mice carrying mutations in multiple genes by CRISPR/Cas-mediated genome engineering. Cell. 2013;153(4):910–8. pmid:23643243
- 4. Aida T, Chiyo K, Usami T, Ishikubo H, Imahashi R, Wada Y, et al. Cloning-free CRISPR/Cas system facilitates functional cassette knock-in in mice. Genome Biol. 2015;16:87. pmid:25924609
- 5. Yoshimi K, Kunihiro Y, Kaneko T, Nagahora H, Voigt B, Mashimo T. ssODN-mediated knock-in with CRISPR-Cas for large genomic regions in zygotes. Nat Commun. 2016;7:10431. pmid:26786405
- 6. Ukai H, Kiyonari H, Ueda HR. Production of knock-in mice in a single generation from embryonic stem cells. Nat Protoc. 2017;12(12):2513–30. pmid:29189772
- 7. Kiyonari H, Kaneko M, Abe S, Aizawa S. Three inhibitors of FGF receptor, ERK, and GSK3 establishes germline-competent embryonic stem cells of C57BL/6N mouse strain with high efficiency and stability. Genesis. 2010;48(5):317–27. pmid:20162675
- 8.
Nagy A, Gertsenstein M, Vintersten K, Behringer R. Manipulating the Mouse Embryo: A Laboratory Manual. Third Edition. New York: Cold Spring Harbor Laboratory Press; 2003.
- 9. Nagy A, Nagy K, Gertsenstein M. Production of mouse chimeras by aggregating pluripotent stem cells with embryos. Method Enzymol. 2010;476:123–49.
- 10. Gertsenstein M, Nutter LM, Reid T, Pereira M, Stanford WL, Rossant J, et al. Efficient generation of germ line transmitting chimeras from C57BL/6N ES cells by aggregation with outbred host embryos. PloS One. 2010;5(6):e11260. pmid:20582321
- 11. Dittrich PS, Manz A. Lab-on-a-chip: microfluidics in drug discovery. Nat Rev Drug Discov. 2006;5(3):210–8. pmid:16518374
- 12. El-Ali J, Sorger PK, Jensen KF. Cells on chips. Nature. 2006;442(7101):403–11. pmid:16871208
- 13. Kovarik ML, Gach PC, Ornoff DM, Wang Y, Balowski J, Farrag L, et al. Micro total analysis systems for cell biology and biochemical assays. Anal Chem. 2012;84(2):516–40. pmid:21967743
- 14. McKim JM Jr.. Building a Tiered Approach to In Vitro Predictive Toxicity Screening: A Focus on Assays with In Vivo Relevance. Comb Chem High T Scr. 2010;13:188–206.
- 15. Abbott A. Cell culture: biology's new dimension. Nature. 2003;424(6951):870–2. pmid:12931155
- 16. Bhatia SN, Ingber DE. Microfluidic organs-on-chips. Nat Biotechnol. 2014;32(8):760–72. pmid:25093883
- 17. Huh D, Torisawa YS, Hamilton GA, Kim HJ, Ingber DE. Microengineered physiological biomimicry: organs-on-chips. Lab Chip. 2012;12(12):2156–64. pmid:22555377
- 18. Wagner I, Materne EM, Brincker S, Sussbier U, Fradrich C, Busek M, et al. A dynamic multi-organ-chip for long-term cultivation and substance testing proven by 3D human liver and skin tissue co-culture. Lab Chip. 2013;13(18):3538–47. pmid:23648632
- 19. Frey O, Misun PM, Fluri DA, Hengstler JG, Hierlemann A. Reconfigurable microfluidic hanging drop network for multi-tissue interaction and analysis. Nat Commun. 2014;5:4250. pmid:24977495
- 20. Rismani Yazdi S, Shadmani A, Burgel SC, Misun PM, Hierlemann A, Frey O. Adding the 'heart' to hanging drop networks for microphysiological multi-tissue experiments. Lab Chip. 2015;15(21):4138–47. pmid:26401602
- 21. Hsiao AY, Torisawa YS, Tung YC, Sud S, Taichman RS, Pienta KJ, et al. Microfluidic system for formation of PC-3 prostate cancer co-culture spheroids. Biomaterials. 2009;30(16):3020–7. pmid:19304321
- 22. Kwapiszewska K, Michalczuk A, Rybka M, Kwapiszewski R, Brzozka Z. A microfluidic-based platform for tumour spheroid culture, monitoring and drug screening. Lab Chip. 2014;14(12):2096–104. pmid:24800721
- 23. Chan HF, Zhang Y, Ho YP, Chiu YL, Jung Y, Leong KW. Rapid formation of multicellular spheroids in double-emulsion droplets with controllable microenvironment. Sci Rep. 2013;3:3462. pmid:24322507
- 24. Abe T, Kiyonari H, Shioi G, Inoue K, Nakao K, Aizawa S, et al. Establishment of conditional reporter mouse lines at ROSA26 locus for live cell imaging. Genesis. 2011;49(7):579–90. pmid:21445964
- 25. Ode KL, Ukai H, Susaki EA, Narumi R, Matsumoto K, Hara J, et al. Knockout-Rescue Embryonic Stem Cell-Derived Mouse Reveals Circadian-Period Control by Quality and Quantity of CRY1. Mol Cell. 2017;65(1):176–90. pmid:28017587
- 26.
Todd RH, Allen DK, Alting L. Manufacturing processes reference guide. 1st ed. New York: Industrial Press; 1994. xxiv, 486 p. p.
- 27.
Lide DR. CRC Handbook of Chemistry and Physics, 83th ed. Boca Raton: CRC Press; 2002.