Nonmammalian Genomic Analysis: A Practical Guide

By Bruce Birren and Eric Lai

Offering detailed protocols for those needing to construct a variety of maps and isolate genes, this unique book is intended to popularize the new techniques of genome analysis derived from the Human Genome Project. The power of these new methods is often most striking when applied to problems outside of human genetics, particularly the nonmammalian systems on which many researchers focus. Many of these organisms are economically important and biologically rich.
Nonmammalian Genomic Analysis: A Practical Guide covers the "how to" aspects of preparation, handling, cloning, and analysis of large DNA and the creation of chromosome and genome maps. This lab manual facilitates the transfer of these technologies to small "low tech" environments and allows them to be used by those with no background in genome mapping or large-fragment cloning. Like having a local expert, this collection provides procedures for anyone, anywhere, and allows the replication of others' success.

• Includes detailed and clearly-written step-by-step protocols
• Evinces expected results and offers trouble shooting advice
• Provides techniques appropriate for small laboratories as well as those with limited resources
• Covers a broad variety of cloning systems, including single copy vectors
• Discusses a diverse range of organisms, from prokaryotes to eukaryotes, from single-celled organisms to highly complex organisms

• Language: English
• Publisher: Academic Press
• Release date: Sep 25, 1996
• ISBN: 9780080537726

Book preview

Germany

Preface

The past decade has seen an explosion in our ability to generate genome maps and isolate genes, with much of this progress being attributable to the concerted effort dubbed the Human Genome Project. However valuable the specific information gathered about the human genome proves to be, the greatest legacy of the Human Genome Project may be that it focused effort on the development of techniques for genome analysis. The power of these techniques is often most striking when they are applied to problems outside human or mammalian genetics. For example, by permitting electrophoretic karyotyping and purification, these techniques have revolutionized studies of chromosome structure, gene mapping, and population biology for species with chromosomes small enough to be resolved by pulsed-field gel electrophoresis. In some cases, the smaller size of nonmammalian genomes means that complete maps may be developed with very little effort or commitment of resources. In the case of organisms with large genomes, these new and efficient methods become essential when genetic maps and mapping resources are lacking. The genomes of some of the world’s economically most important species of animals, plants, and pathogens are still very poorly understood, and limited funding restricts the options of a research laboratory. This book is intended to promote the spread of the new techniques of genome analysis. The procedures selected are those that will be useful to investigators working with a wide variety of organisms, from microorganisms and parasites to complex eucaryotes. Because the examples are all drawn from studies of nonmammalian organisms, there is little need to sift through nonrelevant information, such as approaches specific to human or mouse genetics.

Our goal is to present protocols for procedures exactly as they are carried out in the labs where they are successfully used. Since many of the users of this work will have limited prior exposure to molecular biology, let alone genomic analysis, the procedures are presented in a thorough and detailed fashion. This is important, since failure to successfully transfer many procedures from one lab to another often lies in the details that at first seem routine. Each chapter contains background material to allow novices to understand why each step is used and information on troubleshooting to help identify both when the procedures are working correctly and when they are not, and if so, what corrective measures should be taken. The uniform format permits researchers to rapidly determine both the reagents needed for the work and the steps that must be performed. We expect the book to be used directly at the lab bench by graduate students, postdoctoral scientists, and other researchers carrying out genome analysis, as well as by advanced undergraduates and those directing research efforts who are interested in the strategies and different approaches available for making maps and cloning genes. Although each procedure is illustrated with work involving a single species, all the methods presented will be valuable for study of virtually any genome, and thus we expect each chapter to be of interest to those interested in an overview of the methodology. The book begins with techniques that do not involve cloning, with the first chapter on pulsed-field gels describing methods that underlie virtually all methods described in the rest of the book. The remaining chapters of the book consider different cloning strategies, concluding with an approach to genome mapping that integrates all the kinds of mapping information described in previous chapters.

Bruce Birren and Eric Lai

1

Introduction to Pulsed-Field Gels and Preparation and Analysis of Large DNA

Jennifer S. Lee, Bruce Birren and Eric Lai

I. Pulsed-Field Gel Electrophoresis (PFGE)

A. Introduction

In 1983 Schwartz and his co-workers demonstrated that yeast chromosomes can be separated in agarose gels by using two alternating electric fields of different orientation (Schwartz et al, 1982). This technology is now known as pulsed-field gel electrophoresis (PFGE). PFGE can resolve DNA from a few kilobases (kb) to more than 10 megabases (Mb) long (Orbach et al., 1988), extending the size range of resolution for DNA molecules orders of magnitude beyond conventional agarose electrophoresis. The ability to separate intact chromosomes from microorganisms has revolutionized gene and genome mapping in these species (Carle and Olson, 1985; Fan et al., 1989). Similarly, pulsed-field gel (PFG) techniques have been the basis for many of the advances in large fragment cloning and physical mapping (Shizuya et al., 1992; Dausset et al., 1992). PFGE has become an essential technique for the characterization and analysis of chromosomes and genomic DNA and has recently been reviewed in detail (Birren and Lai, 1993). As both a preparative and an analytical tool, PFGE is central to all aspects of genome analysis, and its use is fundamental to nearly every chapter in this book. In this chapter we review the parameters that govern the migration of DNA in PFGE and provide guidelines and protocols for preparing and manipulating high-molecular-weight DNA.

B. Size-Dependent Separation in PFGE

When DNA in an agarose gel is placed in an electric field, the DNA molecules become elongated and oriented with the electric field as they move through the gel (Smith et al., 1989; Schwartz and Koval, 1989). The leading end represents the head and the following end the tail. In conventional electrophoresis employing a constant electric field, all DNA molecules larger than 15–20 kb travel unidirectionally through the gel with the same mobility. PFGE utilizes two alternately pulsing electric fields of different orientation (Schwartz and Cantor, 1984; Carle and Olson, 1984). With each electric field switch, the DNA reorients to align itself with the new field direction prior to beginning to migrate (Smith et al., 1990). Therefore, the DNA actually migrates in a series of brief steps of alternating direction, with the electric fields regulated to ensure that the net direction of DNA migration is a straight line. The time required for DNA molecules to reorient with each change in field direction is a function of the molecular weight. Thus, by periodically forcing the DNA to reorient, PFGE establishes size-dependent separation for large DNA molecules.

The angle between the two alternating fields is known as the reorientation angle and represents the angle that the DNA must turn through to reorient with each change in field direction. When the two alternating fields are oriented in opposite directions (i.e., a reorientation angle of 180°), the process is referred to as field inversion gel electrophoresis (FIGE) (Carle et al., 1986). In most other cases, the angle between the two fields ranges from 106° to 120°.

C. Switch Time Governs PFG Resolution

PFGs can be run with very little effort or specialized knowledge. However, the researcher must have an adequate understanding of the factors affecting DNA mobility in PFGs to generate reproducible and successful cloning and mapping experiments. Choosing optimal gel conditions in PFGE depends mainly upon the size range of the DNA fragments to be resolved. The goal is to achieve the highest band resolution with minimum run time. Separation of the desired size range of DNA fragments depends on the time required for the molecules to fully reorient from one field direction to the other. The duration of each electric field pulse is referred to as the switch time, or switch interval. For any given switch time, a specific size range of DNA molecules will have had sufficient time to completely reorient and begin to migrate under the influence of the new field. For each switch time, all molecules larger than a certain limit will not have had adequate time to fully reorient with each field switch; these molecules will not resolve from all the DNA in the sample that is equal to or greater than its size. Therefore, the switch time is the single most important parameter in determining which molecules are resolved in a PFG.

Each set of PFGE conditions will separate a specific size range of molecules, but the upper size limit of resolution is primarily determined by the switch time. As the switch time is lengthened, the size range of molecules that can be resolved increases. Figure 1.1 shows that, with all other gel parameters held constant, increasing the switch interval from 45 to 75 sec increases the size range that is separated from 677 to 960 kb, and a further increase to 105 sec permits separation of DNA up to over 1100 kb. Chapter 2 illustrates the importance of careful control of switch time: minor changes in switch time can permit resolution of molecules that are otherwise not separated. However, as the switch time is lengthened, although a larger size range of molecules is being resolved, the resolution between the smaller molecules run on each gel is diminished. This is also illustrated in Fig. 1.1. The separation between the 50- and 100-kb bands of the lambda ladder in the left lane of each panel decreases as the switch time is increased. Therefore, obtaining optimal resolution on a PFG requires use of the minimal switch time necessary to effectively separate the molecules of interest. Using much longer switch times than needed will reduce resolution of the sample bands as well as the molecular size markers, reducing the ability to determine accurate sizes.

Figure 1.1 Switch time determines the resolution and size range separated by a pulsed-field gel. With all other gel parameters held constant, Lambda ladders (left lane) and yeast chromosomes (right lane) were separated with switch intervals of 45, 75, or 105 sec. Each pulsed-field gel can separate a certain size range of DNA molecules, and lengthening the switch time increases this size range. However, as the switch time is lengthened, the separation between the fragments of different sizes (the resolution) diminishes. Gel conditions used were 14 = 9A C, 1% SeaKem LE agarose, 0.5× TBE, 120 = 9A reorientation angle at 6 V/cm.

In all PFGs, regardless of the particular switch time used, there will be a portion of the gel in which DNA molecules migrate in an order that does not reflect increasing size. Some molecules will migrate faster than molecules which are actually smaller, a phenomenon known as band inversion. In most PFGE band inversion occurs only in a small region of the gel, normally near the sample well, and therefore does not interfere with determination of sizes. In a FIGE gel, this reversal of the relationship between mobility and size can occur at either the bottom or the top of the gel and can greatly reduce the amount of gel in which useful size information can be determined. The impact of this phenomenon is minimized by (i) using size markers that extend beyond the range of the molecules being studied, (ii) using size markers that have unique sizes and are easily distinguished, such as yeast chromosomes, in addition to markers with regular spacing, such as ladders, and (iii) analysis of the fragments on gels of several different switch times.

The decreased resolution accompanying longer switch times and the separation of greater size ranges of DNA produce two important practical results. First, obtaining high-resolution separations for molecules of very different sizes will require different PFG runs. For example, the switch time necessary to clearly resolve a 2-Mb DNA molecule will poorly resolve a 200-kb fragment. In this case, two separate PFGs must be run with switch times appropriate for each set of size ranges. Second, accurately determining the size of an uncharacterized molecule usually requires multiple PFGs. The preliminary PFGs need a broad switch time that ensures the fragment will be resolved in order to estimate the size. Subsequent PFGs can use a switch time specific to the observed size range of the fragments which would then provide more accurate size information relative to the size markers.

In Fig. 1.2, DNA mobility in PFGs run with different switch times is shown for DNAs of three different size ranges. The gel conditions that give rise to this migration are shown in Table 1.1, and for Fig. 1.2 a single switch time is used throughout the run duration. Just as the PFGE conditions for DNAs of these different size ranges are very different, the time scales of the three curves are notably different. Switch times can be chosen from Fig. 1.1 by noting the shortest time that permits resolution of the desired size range. For example, Fig. 1.2A shows that a switch time of approximately 8 sec would be appropriate to resolve molecules of 35 kb, provided that the gel is run according to the conditions shown in Table 1.1. Molecules of 35 kb are also resolved using longer switch times, such as those shown in Fig. 1.2B and 1.2C, but the conditions separating larger DNA offer poor resolution for the 35-kb DNA fragments.

Table 1.1

Separation Conditions of Different Size DNA Fragmentsa

aThese are standard conditions for pulsed-field gels, and are chosen to offer the optimal separation in terms of resolution and run time for each of the different size ranges of DNA. Many variations on each of these parameters will also be effective.

bLower voltage gradients (1.5 V/cm) may be necessary for resolving fragments over 4 Mb. Further discussion of separation of Mb DNA can be found in Chapter 2.

cLow EEO agarose refers to EEO values of less than approximately 0.13.

Figure 1.2 Determining the optimum switch time for separating different sized DNA. The sizes of the largest molecules that could be resolved in pulsed-field gels using a constant switch interval are shown for DNAs of three different size ranges. Reproducing the resolution depicted requires use of the gel conditions used to generate each of the three curves, given in Table 1.2. To select an appropriate switch time, i.e., one providing optimal resolution of the desired size range of molecules, one should choose from these curves the shortest time that permits resolution of the desired size range. These curves can be used to predict the separation obtained when using ramped switch times by calculating the average switch time.

Often, the switch time is progressively changed during the run, so that the fields alternate more frequently at the beginning of the run than at the end. This is termed switch time ramping. PFGs that are run with a constant switch time over the entire run duration normally have a specific area of the gel where the fragment mobility is linear with respect to fragment size. Switch time ramping increases the proportion of the gel in which the fragment mobility is linear to its size. This is especially true for FIGE, where failure to ramp the switch time leads to a large portion of the gel that cannot be used due to band inversion. With switch time ramping, the DNA molecules migrate with a mobility that reflects the average of all the switch times used. Thus, the actual size range of molecules separated on a gel with ramped switch times can be predicted by esimating the average switch time between the initial and final value. For example, a gel run using an initial switch time of 20 sec and a final switch time of 120 sec would separate the same approximate size range as a gel run using a constant switch time of 70 sec. While Fig. 1.2 shows the mobility of DNA run with a constant switch time, these curves can also be used to predict the separation obtained when using ramped switch times by calculating the average switch time.

D. Establishing Effective Separation Conditions

While the switch time is the primary determinant of the migration of DNA in PFGE, all of the other conditions under which the gel is run also influence the speed and resolution of the separation (Birren et al., 1988). The greatest changes in DNA migration and resolution occur with switch time variation, followed by changes in the voltage gradient. Most of the parameters that must be selected for a PFG are similar to those involved in conventional electrophoresis: the temperature, the duration of the gel run, the agarose concentration and electroendosmotic (EEO) value, and the buffer type and concentration. The impact of variations in these parameters is much more pronounced in PFGE than in conventional electrophoresises, though these conditions do not greatly change the outcome of a PFG. PFGs also involve the additional electric field parameters, namely the duration and direction of the alternating fields.

Table 1.1 presents conditions that are effective for routine separation of DNA molecules from 1 kb to 10 Mb. Specific applications may benefit from variation of these conditions, for example, to achieve higher resolution over small defined size range or more rapid separations of lower resolution. Depending on the specific parameters changed and/or the combined effect of several parameter changes, some cases will produce minor changes in the migration of the DNA (on the order of 20%), while others can cause a complete failure to resolve the fragments of interest. Therefore, understanding the relationship of these factors to the migration of the DNA allows the researcher to optimize the conditions for any given separation and understand the possible ways to obtain satisfactory separation. When contemplating changes in PFGE conditions, it is important to remember that any change is likely to affect not only the speed of the separation but the resolution as well.

The largest difference between the conditions used for the different DNA size ranges in Table 1.1 is in the voltage gradient. For molecules smaller than 50 kb, voltage gradients as high as 20 V/cm may be used for PFGs (Wagner and Lai, 1994), though commercially available PFGE boxes are limited to gradients of 9 V/cm. These high-voltage gradients produce rapid DNA separations by dramatically increasing the rate of DNA migration. In contrast, with increasing size, DNA molecules require a reduction in the voltage gradient. As described in Chapter 2, separation of Schizosaccharomyces pombe chromosomes (3.5 to 5.7 Mb) can occur only with gradients of 2 V/cm or less (Smith et al., 1987; Vollrath and Davis, 1987). This reduction in voltage gradient necessitates such long gel runs for large DNA that changes in other gel conditions, such as the reorientation angle, become especially important to minimize run durations.

Although a range of temperatures from 4 to 30°C may be used for PFGs, the temperature must be carefully and consistently regulated throughout the PFGE. A pump and usually a heat exchanging system ensures uniform migration and resolution across the lanes of the gel and over the duration of the gel run. PFGE buffers are usually either Tris Borate (TBE) or Tris Acetate (TAE) buffers. TBE is preferred for routine use, since it requires changing less frequently than TAE. For separation of very large DNA, the increased separation speed of TAE buffers is valuable. At the lower voltage gradients used for large DNA PFGE, the buffer will not break down as rapidly as at the higher voltages used for separation of smaller DNA. A reorientation angle of 120° is effective for all PFG separations, though reductions in this angle will allow more rapid separations of DNAs of all sizes.

Table 1.2 summarizes the role of each PFGE parameter and lists the effects of its increase or decrease on the rate of migration of the DNA and the resolution obtained in a PFG. This table can be used to predict the affect of changes in each of these parameters on the PFG. For example, if the gel will be run at temperatures lower than the recommended 14°C, a longer run time is necessary to produce the same degree of separation. More importantly, each factor that influences the rate at which the DNA migrates also affects the size range of molecules that is resolved: each change in a run parameter will alter the size range of molecules that are separated and their resolution. As listed in Table 1.3, otherwise undesirable changes in resolution can be compensated for by concomitant changes in the switch time. In some instances the relationship is easy to predict. When the voltage gradient is reduced by half, the switch time must be approximately doubled to maintain comparable resolution. In most cases, achieving optimum separations requires some amount of trial and error, varying the switch time with most other parameters held constant (for further details see Chapter 2 and Birren and Lai, 1993).

Table 1.2

Parameters That Affect Migration of DNA in Pulsed-Field Gels

Table 1.3

Interaction of Switch Time with Other PFGE Parameters

E. Selection of a PFGE Instrument

Initially, PFGE boxes were home-made instruments involving a variety of designs and names, for which ease of use was not a primary concern (reviewed in Lai et al., 1989). Today, commercial PFGE boxes that are simple to use and offer a range of features are widely available. Nearly all commercially available PFGE instruments are now based on the CHEF (Contour-clamped Homogeneous Electric Field) (Chu et al., 1986) or FIGE (Carle et al., 1986) designs, both of which produce DNA migration in a straight line. The more advanced CHEF systems incorporate the additional features and flexibility of the PACE (Programmable Autonomously Controlled Electrodes) gel box (Clark et al., 1988). Selection of a PFGE instrument depends on the needs of the researcher and the type of separations to be performed. FIGE will adequately separate small DNA fragments (under 200 kb) and has limited use with large DNA, such as generating a large-fragment restriction map around a particular gene or preparing a few blots of separated yeast chromosomes for mapping. FIGE has the advantage that, aside from the switching unit, it requires only standard components such as a gel box and power supply that are usually already present in the lab and can be used after PFGE is no longer needed. FIGE systems are available that use a constant voltage for the forward and reverse fields (such as the Hoefer Switchback). For fragments under 100 kb, superior resolution can be obtained by using a FIGE system that varies the voltage instead of the time (Birren et al., 1989), such as the FIGE Mapper from Bio-Rad. However, for projects involving whole genome mapping, electrophoretic karyotyping, or large fragment cloning, a reliable and flexible PFGE device is essential. The Bio-Rad CHEF DRII and the Pharmacia Gene Navigator are simple fixed-angle (120°) CHEF systems, adequate for separating DNA from 50 kb to 2 Mb if long run times are not a concern. However, features of more advanced systems can noticeably improve PFGE separations and reduce run durations. For example, the Bio-Rad CHEF DRIII has a variable reorientation angle from 106° to 120°. This allows reduction of the reorientation angle which can save days of electrophoresis time for separations of Mb DNA (see Chapter 2). When separating P1, BAC, or PAC digests, the combined use of reorientation angles less than 120° and higher voltage gradients can reduce run durations from 16 to 4 hr (Birren and Lai, 1995). The Bio-Rad CHEF Mapper has the most advanced features, permitting any number of fields to be used with any orientation and duration. In addition, it contains an algorithm that chooses optimal separation conditions based on the size of the DNA molecules of interest. This permits implementation of the most effective separation conditions, specific to each application.

II. Materials

A. Electrophoresis Buffers

PFGE buffers are usually either TBE or TAE. TBE is preferred for routine use, due to its higher buffering capacity; 0.5× TBE buffer can be used for several runs of 30 hr at 6 V/cm without requiring changing or replenishment of the buffer. The buffer should always be changed and the gel box rinsed prior to any preparative electrophoresis. TAE will provide faster migration of the DNA than TBE, and is therefore recommended for larger, slow-moving DNA fragments, though the buffer must be changed with each gel run. Do not include ethidium bromide in the buffer or the gel, because ethidium bromide will slow down the migration of the DNA and alter the size range of fragments separated.

1. 1× TBE

The pH of this mixture will be 8.3.

2. 1× TAE

B. Solutions

1. Bacterial Lysis Solution

This solution may by prepared ahead of time. Add egg white lysozyme to 1 mg/ml final concentration immediately prior to use.

2. Digestion Buffer

Prepare fresh buffer by dissolving sarcosyl in EDTA by shaking. Add proteinase K as a 20 mg/ml stock solution.

3. YPD

YPD is a complete medium for yeast composed of yeast extract, peptone, and dextrose. To prepare 1 liter of YPD mix:

10 g Bacto–yeast extract

20 g Bacto–peptone

Add H2O to bring final volume to 1 liter.

Sterilize by autoclaving. Add 40 ml of 50% glucose.

4. LB

LB is a complete medium for bacteria. To prepare 1 liter of LB mix:

10 g Bacto–tryptone

5 g Bacto–yeast extract

10 g NaCl

Add H2O to bring final volume to 1 liter.

Sterilize by autoclaving.

C. Choice of Agarose

Most PFGs are cast at a 1% concentration using standard agarose sold for DNA electrophoresis. The agarose should be certified for use in molecular biology, because contaminants in impure agarose can degrade DNA or inhibit subsequent enzymatic reactions. Faster DNA migration is obtained with agaroses of low electroendosmosis (EEO), which reflects the internal charge of the agarose. For routine use, low EEO agarose (such as SeaKem LE FMC BioProducts) is effective and inexpensive. The use of medium EEO agarose will reduce the speed of DNA migration by approximately 10–15%. For separating DNA molecules larger than 2 Mb, the very long run times needed can be reduced by using agarose of even lower EEO values (often sold as Pulsed-Field Gel agarose) and/or at a concentration of 0.7%—see Chapter 2 for discussion. For preparation of DNA samples in solid agarose, highly purified and quality tested low-melting agarose (e.g., InCert agarose, FMC BioProducts) is of value only when the samples will be used subsequently for restriction digestion. For preparation of intact chromosomes for separation by electrophoresis, conventional low-melting agarose is effective.

III. Preparation of Nonmammalian Chromosomes

A. General Principles

Traditional techniques for purification of DNA involve organic extraction of proteins and alcohol precipitation. These procedures involve shear forces that will break large DNA fragments to an average size of no more than a few hundred kilobases. At the same time that they developed the electrophoretic techniques to separate large DNA, Schwartz and Cantor (1984) developed methods for purifying megabase-sized DNA in solid agarose to protect the DNA during preparation. Intact cells are mixed with low-percentage low-melting agarose which is then allowed to harden in molds. These solid samples are then treated with enzymes and detergents that will digest the cell wall, membranes, proteins, and other cellular debris, allowing them to diffuse out of the agarose, leaving only the nucleic acid. At its simplest, the treatment can require only proteinase K (or another protease), detergent, and EDTA (used to inhibit endogenous nucleases that could degrade the DNA during the incubations). This is usually the only treatment necessary for organisms that lack a cell wall. Digesting a cell wall usually requires an additional step, to allow access of the cell to the reagents used for DNA isolation. The methods, most often enzymatic, are specific to each organism and vary with differences in the nature of the cell wall.

Detailed protocols for preparing high-molecular-weight DNA from a variety of organisms may be found within the other chapters of this book and elsewhere (Birren and Lai, 1993). While the exact protocol for isolation of large or chromosomal DNA will vary from organism to organism, there are general principles that apply in most cases:

(1) The cells should be from a healthy, actively growing source. Cultures in which growth has ceased often will have undergone thickening of the cell wall, making rapid lysis more difficult, and some amount of cell death, resulting in DNA degradation. When a choice of tissues exists, tissues containing low levels of nucleases should be used for DNA preparation.

(2) Most enzymes are less active in the presence of agarose, and hence cell walls are more efficiently digested with the cells in solution rather than after embedding the cells in agarose. Resulting spheroplasts must be osmotically stabilized during any washing steps to prevent lysis prior to embedding in agarose.

(3) Add EDTA to all solutions as early in the process as possible to prevent nucleolytic DNA degradation. Low-melting agarose should be used because it will remain liquid when cooled to temperatures that will not damage the cells. If the DNA will be digested with restriction enzymes after preparation, the low-melting agarose must be free of compounds that inhibit enzyme activity.

(4) Once cell walls have been digested, rapidly perform any necessary washes and embed the resulting spheroplasts in agarose.

(5) Sufficient time should be allowed to fully digest the embedded cells since residual cellular components can degrade the DNA on storage or interfere with subsequent enzymatic treatment.

(6) Dialysis of the samples after DNA isolation should be extensive enough to remove small DNA fragments as well as the reagents used for DNA preparation, which can severely inhibit enzyme activity.

B. Procedures

Additional procedures for preparation of high-molecular-weight DNA are found in many chapters of this