| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Department of Chemistry & Biochemistry and The School of Medicine, University of South Carolina, Columbia, South Carolina 29208
ABSTRACT
H1t is an H1 histone variant unique to late spermatocytes and early spermatids. Using gene targeting and embryonic stem cell technologies, we have produced mice with a disrupted H1t gene. Homozygous H1t-null mice have normal fertility and show no obvious phenotypic consequence due to the lack of this histone. Biochemical and immunohistochemical approaches were used to show that normal changes in chromosomal proteins occurred during spermatid development, including the appearance and disappearance of transition proteins 1 and 2. Both protamines 1 and 2 are present in normal amounts in sonication-resistant spermatid nuclei from H1t-null mice. Analysis of H1 histones by quantitative gel electrophoresis in enriched populations of pachytene spermatocytes and round spermatids showed that the lack of H1t is only partially compensated for by somatic H1s, so that the chromatin of these cells is H1 deficient. Because H1t is thought to create a less tightly compacted chromatin environment, it may be that H1-deficient chromatin is functionally similar to chromatin with H1t present, at least with respect to permitting spermatogenesis to proceed.
gametogenesis, meiosis, spermatid, spermatogenesis
INTRODUCTION
The formation of a mature mammalian sperm involves the controlled expression of a large number of proteins that are unique to the male germ line [1, 2]. Testis-specific histone variants comprise a well-known set of such proteins [3–5], and H1t is the testis linker histone variant found in late spermatocytes and early spermatids [6–10]. H1 histones help determine the degree of condensation of the fundamental chromosomal fiber (composed of DNA wrapped around nucleosome cores) [11, 12]. In mammals there are six additional H1 variants: the five common H1 variants (H1a–e) [13–15] of somatic cells and H1o, a distinct variant associated with terminally differentiated tissues [16]. H1t is the only variant expressed according to a totally tissue-specific developmental timetable.
H1 histones are not part of the nucleosome core particle, and alternative models exist for the exact binding site of H1 [17–20]. With naked DNA they have a distinct preference for binding sites that offer DNA cross-overs and bind highly cooperatively. But it is not clear how this behavior relates to binding to nucleosomal DNA. Because highly compacted chromatin is probably inaccessible for transcription or replication, it is appealing to suppose that different H1 variants permit the cell to alter the tightness of chromatin condensation in a regional or developmental fashion. The large number of sequence changes that distinguish H1t from other H1 variants [4, 21–25] suggests that H1t has a particular functional role. It is released from chromatin more readily by salt than are the standard H1s [26] and in vitro is less effective at condensing nucleosomal DNA than other H1 variants [27–29]. Accordingly, it is believed that H1t imparts a less tightly condensed chromatin state. In the rat H1t can account for as much as 55% of the total H1 in late pachytene spermatocytes and round spermatids [7].
The functional rationale for having testis-specific histone variants, and for the existence of H1t in particular, is not known. H1t appears at substantial levels relatively late in the first meiotic prophase, well after chromosomes have undergone pairing [7–10, 30]. It could facilitate the final stages of meiosis but is clearly not necessary for pairing itself and is not present during meiosis in the female [31]. Following meiosis in the male, there is extensive tissue-specific gene expression in early haploid cells [1]. Later, a massive structural reorganization of the nucleus occurs. Histones are replaced stepwise by a set of nuclear transition proteins (of which TP1 and TP2 are most prominent), which are in turn replaced by the protamines that are characteristic of the mature sperm nucleus [3, 9, 32, 33]. Because formation of the condensed sperm nucleus is specific to the male gamete, one of the roles of H1t might be to facilitate the initial histone to transition protein changeover that occurs as the spermatid nucleus undergoes elongation and compaction.
Hoping to gain insights into the functional role of H1t, we have engineered a targeted mutation in the H1t gene in mice. Mice lacking H1t are fully fertile and have no obvious defect in spermatogenesis. However, failure to produce H1t is not completely compensated for, so that late spermatocytes and round spermatids of the H1t-null mice have H1-deficient chromatin. While this work was being completed, an independent report of an H1t knockout mouse appeared by Lin et al. [34].
MATERIALS AND METHODS
Generation of Mice with a Disrupted H1t Allele
The H1t gene was isolated from a 129SvJ mouse liver genomic DNA library kindly provided by T. Doetschman. A targeting vector was constructed from a 10.7-kilobase (kb) EcoRI fragment by deleting a 529-base pair (bp) BssHII-SmaI fragment, eliminating a region extending from just before the TATA box to codon 117 of the coding region. A PGKneo expression cassette was inserted at the site of this deletion, and pMC1-TK cassettes were added at either end of the targeting unit (Fig. 1). Following introduction of the linearized targeting vector into embryonic stem (ES) cell line E14TG2a [35] by electroporation [36], G418-, gancyclovir-resistant colonies were expanded and examined for homologous recombination by Southern blotting (Fig. 1, probe 1). Four targeted clones were identified and used to inject into C57BL/6 blastocysts to produce chimeric mice [36]. Chimeras were bred to C57BL/6 mice, and germline transmission was established for two lines that had identical phenotypes. The mutation was maintained on a mixed 129/C57B6 genetic background. CD-1 mice used in fertility tests were obtained from Harlan Sprague-Dawley (Indianapolis, IN).
|
Histology and Immunohistochemistry
Testes were fixed by immersion in Bouins fixative for 24 h, embedded in paraffin, sectioned, and stained with hematoxylin and eosin. Bouins-fixed tissue was immunostained with anti-H1t [9] and anti-TP2 [37]. Tissue for anti-TP1 staining [38] was fixed in ethanol-acetic acid (3:1) in the cold overnight.
Cell Separation and Histone Isolation
Decapsulated testes were digested sequentially with collagenase and trypsin to produce a single-cell suspension of primarily germ cells as described [39]. The cell suspension was fractionated to enrich for round spermatids and pachytene spermatocytes by centrifugal elutriation [39] with rotor speeds and flow rates essentially as described by Grabske et al. [40] with a Beckman J-21C centrifuge and JE6B rotor. In a typical run, 2 x 108 mixed germ cells from three adult mice were loaded into the rotor, and 6 x 107 round spermatids of about 85% purity were eluted at 2000 rpm and 14 ml/min flow rate. Pachytene spermatocytes (3 x 107) of about 75% purity were eluted at 2000 rpm and 30 ml/min flow rate. Purity was judged by nuclear size and chromatin staining using methanol-acetic acid (3:1) fixation and Giemsa stain. Total histones were extracted from cell fractions as described by Brown et al. [41].
Polyacrylamide Gel Electrophoretic Separation of Chromosomal Proteins
Tissues were homogenized in 4 volumes of ice-cold 5% perchloric acid. After centrifugation at full speed in a microcentrifuge, the supernatant was adjusted to 20% trichloroacetic acid (TCA) from a 100% stock. After centrifugation, the pellet was washed with acetone, dried, and dissolved in 0.1% acetic acid.
Two-dimensional gels were run essentially as described previously [6], with the first dimension in the pH 4.5 system of Reisfeld et al. [42] and the second dimension in a 15% acrylamide SDS slab [43].
Single-dimension SDS slabs were stained with Coomassie blue R250 and digitized using a Umax flatbed scanner in transmissive mode and Adobe Photoshop (Adobe Systems, San Jose, CA) software. The green channel was used for quantitation with NIH Image software (available on the Internet: http://rsb.info.nih.gov/nih-image/) in the uncalibrated optical density mode, following background subtraction. A linear relationship between band intensity and applied histone samples was verified, and experimental samples were adjusted to lie within the linear range. Statistical significance was calculated by Students two-tailed t-test.
Preparation and Analysis of Basic Proteins from Sonication-Resistant Nuclei
Sonication-resistant nuclei [44] were prepared from testes homogenized in 5 ml of ice-cold 0.32 M sucrose, 5 mM MgCl2, 10 mM Tris-HCl, pH 7.5, 0.5 mM PMSF, and suspended in 5 ml of 10 mM Tris-HCl, pH 7.5, 0.2% 2-mercaptoethanol. Soluble proteins extracted by 0.5 M HCl were treated with TCA to yield 4%-insoluble and 4%-soluble fractions. These samples were analyzed in a minislab gel containing 5% (v/v) acetic acid, 2.5 M urea, 15% acrylamide, 0.15% methylenebisacrylamide [45].
RESULTS
Spermatogenesis Proceeds Normally in the Absence of H1t
The H1t gene from strain 129 mice was isolated from a genomic library, and a 1.7-kb XbaI restriction fragment encompassing the gene was sequenced (Fig. 1). As is typical for replication-dependent histone genes, H1t lacks introns. The gene was disrupted by deleting a 519-bp restriction fragment that included the transcriptional start site as well as the amino-terminal and central globular domains of the protein. A neo expression cassette was inserted at the site of the deletion (Fig. 1). A 10.7-kb mouse genomic fragment with the disruption was inserted between a pair of thymidine kinase expression cassettes, and the linearized targeting construct was introduced into mouse ES cells by electroporation. Clones in which homologous integration had occurred were identified by Southern blotting using a downstream flanking probe (Fig. 1). Correctly targeted cells were used to generate chimeric mice that were bred to establish germline transmission of the disrupted allele. Heterozygotes were subsequently cross bred to generate homozygotes.
Testis histology for the H1t-null animals was normal, with no indication of an elevation in the number of degenerating germ cells (Fig. 2, A–D). Meiosis and the development of condensed nuclei in spermatids occurred normally, as judged by light microscopy. H1 histones were extracted from testes of +/+ and -/- mice and analyzed by two-dimensional gel electrophoresis for the presence of H1t (Fig. 3). No traces of H1t were observed in the H1t-null animals (Fig. 3C) in contrast to wild-type littermates (Fig. 3B). Separation of thymus H1s on the same gel served to emphasize the unusual H1 composition of the testis, in which H1a, H1c, and H1t are normally the major variants (Fig. 3A). As further demonstration of the absence of H1t in the H1t -/- mice, we compared the immunoreactivity of testis cross sections from +/+ and -/- mice to anti-H1t. While H1t immunoreactivity in normal animals is first detected in late pachytene nuclei, the immunostaining is most prominent in round and early elongating spermatids, perhaps because the chromatin is less condensed than in the spermatocytes (Fig. 2, E and F). No immunoreactivity was seen with sections from H1t-null mice, as expected (Fig. 2, G and H).
|
|
Mice Lacking H1t Are Fertile
To examine the fertility of H1t-null males, five 2- to 9-mo-old animals were bred to CD1 females in parallel with normal littermates. Each male sired a litter of 14 or 15 pups within 3–4 wk of being put with the females, regardless of genotype. From this result we concluded that the H1t-null males have normal fertility. Examination of testis homogenates from these mice for condensed, sonication-resistant spermatid nuclei showed that there was no significant difference between H1t-null animals (3.9 x 107 ± 0.33 SD, n = 5) and their (+/+) littermates (4.0 x 107 ± 0.29 SD, n = 5) spermatid nuclei/testis. Microscopic examination of epididymal sperm under phase contrast indicated no obvious differences in morphology due to lack of H1t (results not shown).
Nuclear Transition Proteins and Protamines in H1t-Null Mice
One of the major developmental changes of spermatids is the conversion of chromatin from a histone-associated, nucleosomal arrangement, to the highly compacted, protamine-associated state in the mature sperm head [46]. This changeover occurs by a program in which histones are first replaced by TP2 and TP1, with protamines appearing only later. Because a role of testis-specific histone variants could be to facilitate the histone-to-transition protein part of this progression, It was of considerable interest to determine if the normal pattern of chromosomal protein changes was occurring in spermatids of the H1t-null mice.
The appearance of the major transition proteins is associated with resistance of spermatid nuclei to disruption by sonication [44], and this property is retained as protamines eventually replace the transition proteins. Accordingly, analysis of the acid-soluble proteins of sonication-resistant nuclei is a useful way to estimate the levels of transition proteins as well as protamines. We prepared sonication-resistant nuclei and examined the 0.5 M HCl soluble proteins following differential precipitation with TCA. No reproducible differences were observed in these extracts from +/+ and -/- mice (Fig. 4). Amounts of TP2 and TP1 (found primarily in the 4–20% TCA fraction) were comparable. Protamine formation in the mouse is complex, because protamine 2 is made as a precursor and converted into mature protamine 2 by a series of proteolytic steps [47, 48]. Thus, several precursor forms are present in significant quantities in addition to mature protamine 2. In contrast, protamine 1 is not made as a precursor. Protamine 1 is found exclusively in the 0–4% TCA fraction, whereas, protamine 2 and its precursors partition about equally into both the 0–4% and 4–20% TCA fractions. Examination of the various protamine components also failed to show any significant differences between +/+ and -/- mice. Thus there were no quantitative differences in the expression of any of these proteins of maturing spermatids that could be attributed to the absence of H1t.
|
To rule out more subtle changes, such as a delay in the initial appearance of TP1 or TP2, we carried out immunohistochemical localization of each transition protein on testis cross sections. We noticed no differences between the pattern of immunoreactivity For TP2 (Fig. 2, I–L) or for TP1 (Fig. 2, M–P) between H1t-null and wild-type mice. Thus, the presence of H1t is not obligatory for the orderly changes in structural chromosomal proteins that occur in spermatids.
H1t-Null Mice Are H1-Deficient in Late Spermatocytes and Round Spermatids
An important remaining question was whether the lack of H1t led to an H1-deficient chromatin state in the H1t-null mice. To address this issue, we isolated cell populations enriched for round spermatids and pachytene spermatocytes using enzymatic digestion of testis tubules and centrifugal elutriation. Total histones were extracted and resolved on SDS slab gels. Stained bands of H1 variants and H4 were scanned and quantitated, and the ratios of H1 fractions to H4 were determined. With both the spermatocytes and round spermatids, the total H1 content was significantly reduced from the H1t-null animals, because the absence of H1t was only partially compensated by increases in other variants (Fig. 5). While the fraction of somatic-type H1 components was somewhat elevated in the null animals, the increase was not sufficient to offset the absence of H1t. The H1 complement of the spermatid fraction was about 74% of wild type, and the H1 complement of the spermatocyte fraction was 79% of wild type. Because these estimates are based on absorbance values taken from stained gels, there is some uncertainty as to whether H1t has the same affinity for Coomassie blue as the standard H1 variants. However, the extinction coefficient for the H1t-dye complex would have to be at least twice as great as for the standard H1 variants to explain away the differences we obtained. Because H1t is somewhat shorter than other H1s (207 residues for H1t vs. 211–212 for mouse H1a and c), it is possible that H1t is slightly underestimated by this procedure, so that the actual difference between mutants and wild type is slightly greater and not less than our estimates. Furthermore, because there is generally expected to be one molecule of H1 for each pair of H4, the total H1:H4 absorbance ratio of about 1, obtained with wild-type mice agrees with this expectation, recalling that a molecule of H1 is just over twice the length of a molecule of H4 (102 residues).
|
DISCUSSION
Using ES cell technology, we generated a targeted mutation in the mouse H1t gene and investigated the phenotype of mice that were homozygous for the disrupted H1t allele. The null mutation had no apparent effect on spermatogenesis or on fertility. Testis histology was normal as were the numbers of condensed spermatid nuclei per testis. Mutant mice readily sired large litters from CD-1 females. Levels of the major transition proteins TP1 and TP2 as well as protamines 1 and 2 were indistinguishable from those of normal mice. The developmental appearance of TP1 and TP2 in the nuclei of condensing spermatids was normal, as judged from immunohistochemical detection. In short, there was no obvious phenotypic consequence for spermatogenesis due to the complete absence of testicular histone variant H1t.
This result was unexpected in view of the striking difference in sequence between H1t and the other common somatic variants [4, 21–25] and in light of the substantial concentrations of H1t in late spermatocytes and early spermatids [7]. H1t is now believed to impart a relatively noncondensed state to chromatin based on in vitro studies of its poor ability to condense or aggregate H1-depleted nucleosome arrays [27–29]. While the details of how H1 binds to the nucleosome are not yet known with certainty, crystal studies of the globular part of H5 (an H1 relative found in the nucleated red blood cells of birds and fish) led to the prediction of two discrete DNA binding sites on the globular region, the second of which was dependent on a cluster of four basic residues that are extremely conserved [17]. Interestingly, H1t lacks one of these basic residues, having a glutamine at position 53 instead of the lysine that is usually present at this position (homologous to lys40 of goose H5). Mutational alteration of all the residues in this second site affected the ability of H5 to assemble cooperatively in complexes containing two DNA duplexes [49]. In a similar mutational analysis of H1o, Hayes et al. [50] found that change of just the single basic residue corresponding to position 53 of H1t led to the virtual loss of ability of H1o to form aggregates with naked DNA. While mutational studies of the role of residue 53 in H1t have not been reported, one can speculate that the absence of this basic residue could contribute to the poor chromatin-condensing ability of H1t.
Because of the male-specific expression of H1t [31], speculation as to its role tends to focus on male-specific features of gamete formation, such as the haploid gene expression that occurs in spermatids [1] and the extensive nuclear reorganization that occurs beginning midway in spermatid differentiation [3]. We have not examined H1t-null mice for subtle changes in gene expression in spermatids. Rapid progress in DNA array technology may make such analysis feasible in the near future. A general relaxation of chromatin higher ordered structure seems to fit well with the fact that a massive reorganization of chromosomal proteins will occur shortly. The absence of H1t did not alter this reorganization as far as we could detect. However, our analysis of purified cell populations indicated that failure to make H1t left the chromatin of spermatids with only 75% of the total H1 normally present. It seems possible that this H1 deficiency could also facilitate access to chromatin by the transition proteins and so lead to the normal conversion of nucleosomal DNA to protamine-condensed DNA. In their independent study of an H1t-null mouse, Lin et al. [34] also failed to observe any effect on spermatogenesis, but these authors reported that other H1 variants fully compensated for the loss of H1t. This discrepancy may be accounted for by the fact that we analyzed enriched populations of the germinal cells that contain H1t, while Lin et al. analyzed only total germinal cells.
If H1t-deficient chromatin is functionally as good as H1t-associated chromatin, what then is the function of H1t? Obviously, this can only be a subject of speculation at present. One factor to consider is whether H1-deficient chromatin is more or less susceptible to DNA damage. Of various mutagens that are known to function in the postmeiotic period of spermatogenesis, some, such as chlorambucil, act in round spermatids [51–53]. It would be interesting to know if such mutagens are more or less effective in H1t-null mice. Recently, the hypothesis has been advanced that a major role of methionine in proteins is to provide a repairable site to scavenge reactive oxygen species [54]. H1t is remarkable among H1 histones for its high methionine content, because the typical H1 contains no methionine at all. It may be possible to design experiments to test if DNA is more susceptible to oxidative damage in round spermatids from H1t-null mice than from normal animals.
The demonstration of specific functional roles for H1 variants through mutagenic elimination has in general not revealed the expected phenotypic effects. Tetrahymena that lack H1 altogether in their functional nucleus are not obviously defective, despite having less condensed chromatin [55], although some subtle changes in gene expression were also observed [56]. Elimination of all the H1 variants but one from chicken cells in culture did not effect their growth and had only minor changes on protein profiles [57]. Targeted elimination of H1o [58] and H1a [59] in mice has also had no readily detectable effect.
In view of our results, an informative experiment would be to produce a transgenic mouse line from H1t-null animals in which the H1t control region was directing production of one of the standard H1 variants that is not common in spermatids, such as H1d or H1e. Were H1d to substitute for H1t, the normal H1 content might be maintained and any phenotype would be due to absence of H1t rather than a change in overall H1 content. Obviously there remains much to do in order to understand the biological rationale for the diversity of linker histones.
NOTE ADDED IN PROOF
Drabent et al. [60] have also reported formation of mice that lack Hlt.
ACKNOWLEDGMENTS
Electroporation of ES cells and production of chimeric mice was carried out in a facility under the direction of Beverly Koller at the University of North Carolina. We thank Anne Latour of the Koller lab for excellent assistance in this project, Dimitri Spyropoulos for advice in the construction of the targeting vector and for a plasmid with the dual thymidine kinase expression cassettes, Tom Doetschman for the genomic library, Michael Rudnicki for the PGKneo cassette, and Rich Showman for use of his microscope.
FOOTNOTES
First decision: 25 August 2000.
1 This work was supported by NIH grant HD-10793. 
2 Correspondence: W.S. Kistler, Department of Chemistry and Biochemistry, University of South Carolina, 631 Sumter St., Columbia, SC 29208. FAX: 803 777 9521; kistler;camail.chem.sc.edu
3 Current address: Department of Molecular Biology and Pharmacology, Washington University School of Medicine, St. Louis, MO 63110.
4 Current address: Merck, Overland Park, KS 66210.
Accepted: September 11, 2000.
Received: July 20, 2000.
REFERENCES
chain gene. Cell 1993; 75:969–976.[CrossRef][Medline]This article has been cited by other articles:
|
|
 |
|
  P. Grzmil, D. Boinska, K. C. Kleene, I. Adham, G. Schluter, M. Kamper, B. Buyandelger, A. Meinhardt, S. Wolf, and W. Engel Prm3, the Fourth Gene in the Mouse Protamine Gene Cluster, Encodes a Conserved Acidic Protein That Affects Sperm Motility Biol Reprod, June 1, 2008; 78(6): 958 - 967. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Martianov, S. Brancorsini, R. Catena, A. Gansmuller, N. Kotaja, M. Parvinen, P. Sassone-Corsi, and I. Davidson Polar nuclear localization of H1T2, a histone H1 variant, required for spermatid elongation and DNA condensation during spermiogenesis PNAS, February 22, 2005; 102(8): 2808 - 2813. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. C. Horvath, W. S. Kistler, and M. K. Kistler RFX2 Is a Potential Transcriptional Regulatory Factor for Histone H1t and Other Genes Expressed During the Meiotic Phase of Spermatogenesis Biol Reprod, November 1, 2004; 71(5): 1551 - 1559. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Q. Lin, A. Inselman, X. Han, H. Xu, W. Zhang, M. A. Handel, and A. I. Skoultchi Reductions in Linker Histone Levels Are Tolerated in Developing Spermatocytes but Cause Changes in Specific Gene Expression J. Biol. Chem., May 28, 2004; 279(22): 23525 - 23535. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Yan, L. Ma, K. H. Burns, and M. M. Matzuk HILS1 is a spermatid-specific linker histone H1-like protein implicated in chromatin remodeling during mammalian spermiogenesis PNAS, September 2, 2003; 100(18): 10546 - 10551. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Fan, T. Nikitina, E. M. Morin-Kensicki, J. Zhao, T. R. Magnuson, C. L. Woodcock, and A. I. Skoultchi H1 Linker Histones Are Essential for Mouse Development and Affect Nucleosome Spacing In Vivo Mol. Cell. Biol., July 1, 2003; 23(13): 4559 - 4572. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. O. Zalensky, J. S. Siino, A. A. Gineitis, I. A. Zalenskaya, N. V. Tomilin, P. Yau, and E. M. Bradbury Human Testis/Sperm-specific Histone H2B (hTSH2B). MOLECULAR CLONING AND CHARACTERIZATION J. Biol. Chem., November 1, 2002; 277(45): 43474 - 43480. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Xue, H. A. Tarnasky, D. E. Rancourt, and F. A. van der Hoorn Targeted Disruption of the Testicular SPAG5/Deepest Protein Does Not Affect Spermatogenesis or Fertility Mol. Cell. Biol., April 1, 2002; 22(7): 1993 - 1997. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Fan, A. Sirotkin, R. G. Russell, J. Ayala, and A. I. Skoultchi Individual Somatic H1 Subtypes Are Dispensable for Mouse Development Even in Mice Lacking the H10 Replacement Subtype Mol. Cell. Biol., December 1, 2001; 21(23): 7933 - 7943. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. C. Horvath, S. E. Clare, M. K. Kistler, and W. Stephen Kistler Characterization of the H1t Promoter: Role of Conserved Histone 1 AC and TG Elements and Dominance of the Cap-Proximal Silencer Biol Reprod, October 1, 2001; 65(4): 1074 - 1081. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
