• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Nature. Author manuscript; available in PMC Feb 21, 2009.
Published in final edited form as:
PMCID: PMC2597896
NIHMSID: NIHMS73506

Molecular basis of the copulatory plug polymorphism in C. elegans

Abstract

Heritable variation is the raw material for evolutionary change, and understanding its genetic basis is one of the central problems in modern biology. We investigated the genetic basis of a classic phenotypic dimorphism in the nematode Caenorhabditis elegans. Males from many natural isolates deposit a copulatory plug after mating, whereas males from other natural isolates—including the standard wild-type strain (N2 Bristol) that is used in most research laboratories—do not deposit plugs1. The copulatory plug is a gelatinous mass that covers the hermaphrodite vulva, and its deposition decreases the mating success of subsequent males2. We show that the plugging polymorphism results from the insertion of a retrotransposon into an exon of a novel mucin-like gene, plg-1, whose product is a major structural component of the copulatory plug. The gene is expressed in a subset of secretory cells of the male somatic gonad, and its loss has no evident effects beyond the loss of male mate-guarding. Although C. elegans descends from an obligate-outcrossing, male-female ancestor3, 4, it occurs primarily as self-fertilizing hermaphrodites57. The reduced selection on male-male competition associated with the origin of hermaphroditism may have permitted the global spread of a loss-of-function mutation with restricted pleiotropy.

Keywords: natural variation, mucin, copulatory plug, C. elegans, plg-1

The synthesis of evolutionary and developmental genetics demands empirical data on the structure of heritable variation in populations. The required data include the identities of the genes that underlie variation, the mutations that affect those genes, and the positions of the genes within the networks of molecular interactions that shape phenotypes. C. elegans is a promising system for the discovery of the molecular basis of variation. Although most studies of C. elegans have employed a single genetic background, natural isolates exhibit a wide range of phenotypes1, 8. Some of this phenotypic variation may result from the change in selective environment associated with the recent origin of hermaphroditism in C. elegans. Males arise through rare non-disjunction events and contribute at a low rate to outcrossing57. Because their relatives in the Caenorhabditis clade produce copulatory plugs, the plugging strains of C. elegans likely retain an ancestral trait that confers significant advantages in male-male competition for matings (supplementary information and Figure S1). The failure of some strains of C. elegans to produce plugs may represent a loss-of-function mutation with minimal impact on fitness due to the diminished competition among males. Hodgkin and Doniach1 showed that plugging exhibits Mendelian inheritance attributable to a single locus, plg-1. The plugging phenotype is dominant, consistent with a model of evolutionary and molecular loss of function in the N2 strain. We tested this model and characterized the molecular basis for heritable variation in copulatory plugging by identifying the gene and molecular lesion underlying plg-1.

Hodgkin and Doniach localized plg-1 to a large region of chromosome III. We employed genetic mapping in crosses between the non-plugging strain N2 and a plugging natural isolate from Palo Alto, CB4855, to refine the location of plg-1 to a 74-kb interval (supplementary methods and Figures S2-S3). Independently, we mapped plg-1 to a coincident interval in a panel of recombinant inbred lines between N2 and a plugging natural isolate from Hawaii, CB4856 (Figure S4). A search for PCR-product length polymorphism showed that a retrotransposon present in the interval in the non-plugging strain N2 is absent in the plugging strains. Sequence analysis revealed that the retrotransposon and its long terminal repeats (LTRs) interrupt a novel, unannotated protein-coding gene whose predicted product has similarity to canonical mucins (Fig. 1a). Mucins are large glycoproteins that are often secreted to form gelatinous substances, including mucus, in diverse metazoans9; a mucin is therefore a natural candidate for a major structural component of the plug. The protein is predicted to contain proline-, threonine-, and serine-rich (PTS) repeats, characteristic of highly O-glyclosylated mucoproteins10, 11 (Fig. 1b); each PTS repeat contains 15 predicted O-glycosylation sites12. The non-repetitive N- and C-termini of the predicted protein are evolutionarily conserved among Caenorhabditis species, and the N-terminus includes a predicted 21 amino acid secretory signal peptide13 (Fig. 1c). Each class of PTS repeats is much more similar within species than between species, consistent with standard models of concerted evolution of repetitive sequences14 (Fig. 1d).

Figure 1
The plg-1 gene

We used RT-PCR to show that the stable transcript of the putative mucin is absent in non-plugging strains. A spliced transcript from the gene is present in RNA from adult males of plugging strains, but undetectable in adult males of non-plugging strains (Figure S5). The transcript is also undetectable in adult hermaphrodites from plugging strains, consistent with a male-specific function.

We used two functional tests to confirm that the disruption of the putative mucin is responsible for the loss of plugging. First, we knocked down the expression of the gene with RNAi in a plugging strain. The resulting males failed to produce copulatory plugs, confirming that the gene is necessary for copulatory plug formation (Fig. 2a). Second, we introduced a transgene carrying the putative mucin into a non-plugging strain. This was sufficient to transform the strain into one that produced plugs (Fig. 2b). We conclude that the putative mucin is the gene underlying the plg-1 locus, and the absence of the disruptive retrotransposon is the plugging allele e2001 defined by Hodgkin and Doniach1.

Figure 2
The mucin gene is plg-1

We next confirmed that the copulatory plug consists of an O-glycosylated mucin. We isolated hermaphrodites immediately after mating, and stained the resulting copulatory plugs with a fluorescent-tagged peanut agglutinin lectin that recognizes Galβ1,3GalNAc-S/T, a carbohydrate modification characteristic of mucins15, 16. The lectin stained the plugs specifically (Fig. 3a–b), indicating that an important constituent of each plug contains a carbohydrate structure that is typical of mucin glycoproteins.

Figure 3
The copulatory plug contains carbohydrate structures typical of mucins and plg-1 is expressed in only a small number of male cells

We examined expression of plg-1 by introducing a transgenic construct with the evolutionarily conserved 5′ upstream region of plg-1 driving GFP. We observed strong expression of GFP in twelve cells of the vas deferens (Fig. 3c–f), a tube of 30 secretory cells connecting the seminal vesicle to the cloaca17. Expression commences during the fourth larval stage, prior to the adult molt. In adult males, the cuboidal GFP-expressing cells are arranged in two rows (Fig. 3c). There is left-right asymmetry in the location of the GFP-expressing cells and in the intensity of their expression. Although there appears to be variation among individuals, typically five of the cells are on the right-hand side of the midline, whereas seven are on the left-hand side of the midline; furthermore, three posterior cells express a lower level of GFP, and two of these low-level cells are on the left-hand side (Supplementary Movie 1).

plg-1::GFP provides the first marker for this suite of vas deferens cells, a previously uncharacterized feature of male anatomy, and reveals their dynamic role in mating. Unmated males have a single large vacuole that occupies most of the cell volume in each of the twelve cells (Fig. 3d–e); in contrast, males examined immediately after mating lack these large vacuoles, and the twelve cells exhibit instead multiple smaller vacuoles (Fig. 3f). The vacuoles are present in both plugging and non-plugging genotypes and are likely storage compartments for seminal fluid.

We next assessed the geographic distribution of plg-1 alleles. We tested 153 natural isolates from around the globe for the ability to complement the N2 loss-of-function allele. Forty-eight strains from Europe, North America, and Australia failed to complement N2, and all carried the retrotransposon insertion in plg-1. All 105 strains that complemented N2 lacked the insertion (Fig. 4). As our PCR assay confirms that the location of the insertion and its orientation is shared among the non-plugging strains (supplementary information), we infer that the common loss-of-function allele of plg-1 represents a single mutation that has spread across the globe. Plugging and non-plugging strains co-occur within localities, and the non-plugging genotype persists across years5, 6, suggesting that the loss of plugging incurs little fitness cost in nature.

Figure 4
The plg-1 loss-of-function allele is globally distributed and Cer1 has been recently active

In the complete genome sequence of N2, the insertion of Cer1 into plg-1 is the sole intact copy of the retrotransposon18. We assayed natural isolates of C. elegans with uninterrupted plg-1 sequences by PCR for the presence of Cer1 elsewhere in the genome (Table S1). While 25 strains had no evidence of Cer1 or its LTRs, 59 strains carry the retrotransposon and its LTRs and another 21 carry the LTRs alone, evidence of having had and later lost the retrotransposon. We found by sequence analysis that Cer1, a gypsy/Ty3 type retrotransposon19, is similar to retrotransposons in the genome sequences of C. briggsae, C. brenneri, and C. remanei, and these mobile genetic elements may be a common source of new mutations in populations of Caenorhabditis.

Induced mutations in a large number of genes disrupt male mating and copulatory plug production2022, but the naturally occurring allele of plg-1 pinpoints the terminal gene in the plug-production network, a structural component of the plug itself. Genes upstream of plg-1 in the male mating network, including genes required for cilium development, neurotransmitter biosynthesis and transmission, and male tail morphogenesis, have extensive pleiotropic roles in C. elegans biology22. In contrast, the plg-1 gene has no apparent roles outside of plugging. We found that the gene is expressed exclusively in males in a subset of cells with no other known function. Males with the loss-of-function allele of plg-1 suffer no evident consequences aside from the loss of plugging. This result is consistent with a role for pleiotropy in filtering the mutations that actually contribute to evolutionary change from the much larger set of mutations that possibly could, a principle laid out by Fisher nearly 80 years ago23. Mutations with the fewest effects are those least likely to change some traits for the worse and hence are those most likely to contribute to evolutionary divergence.24

The only other naturally occurring allele known to disrupt male mating is a loss-of-function allele of mab-23, a gene required for development of the male copulatory organs, and this allele is known only from a single natural isolate1. Although mab-23 is expressed in both males and hermaphrodites, it exhibits an obvious mutant phenotype only in males25, consistent with restricted pleiotropy; at the same time, the rarity of the mab-23 allele suggests that complete loss of male function might be deleterious. Although C. elegans reproduces primarily by self-fertilization in hermaphrodites, selection likely acts to preserve male function2628, eliminating mutations that disrupt male development, morphology or behavior. The high frequency of the plg-1 loss-of-function allele, its co-occurrence with the functional plg-1 allele, and its temporal persistence all suggest that the frequency of males in typical populations is insufficient to select for mate guarding, which requires not just functional males but competition among males. The evolution of hermaphroditism in C. elegans appears to have altered the selective regime of genes required for male-male competitive function. While it is possible that the loss of plg-1 is selectively favored, it may also represent a neutral change drifting to high frequency in the absence of selection against it.

Methods Summary

Nematode Stocks

Nematodes were cultured using standard protocols30.

Sequence Analysis

We merged partial gene predictions across Cer1 in the N2 genome (www.wormbase.org) and then refined the gene model using RACE with RNA from CB4855 males (Clontech SMART RACE kit). Sequences from other species were collected using synteny tracks (www.ucsc.genome.org); we identified N- and C-terminal plg-1 sequences based on sequence conservation and generated alignments of flanking and intervening sequence using LAGAN29.

Expression Analyses

RT-PCR primers flank intron three of plg-1 and fall downstream of the Cer1 insertion site; chimeric Cer1:plg-1 transcript should be detected if present. We generated integrated plg-1::GFP transgene (allele qqIs1) by cloning 311 bp of 5′ upstream sequence into pSM_GFP and cobombarding it with unc-119(+) plasmid pMM016b into QX1015, which carries unc-119(ed3) and plg-1(N2) introgressed into CB4856. We imaged the GFP using a Zeiss 510 META confocal microscope.

Functional Confirmation Experiments

We cloned 3 kb of CB4856 plg-1 exon 4 into RNAi feeding vector L4440 and fed worms E. coli HT115(DE3) carrying the vector or L4440 with no insert. Transgenic rescue of plugging employed a construct derived from CB4856 spanning from 686 bp 5′ of the RACE-defined start of transcription to 178 bp 3′ of the RACE-defined 3′ UTR.

Mucin Staining

Virgin adults were cleaned of bacteria in PBS, allowed to mate for 30 min on clean agarose plates, then fixed and stained with fluorescently labeled lectin. A labeled antibody that fluoresces at a different wavelength but does not bind to nematodes was included to control for the possibility that the gelatinous plug absorbs probe indiscriminately.

Natural Isolates

Each isolate was mated to N2 males and the F1 males tested individually in overnight assays with single N2 hermaphrodites for the ability to produce plugs.

Methods

RT-PCR assays

RNA was isloated using the RNeasy Mini Kit (Qiagen) and amplified using the OneStep RT-PCR Kit (Qiagen) with primers specific for plg-1 (plgRT125F and plgRT259R, flanking intron 3) and for the male-expressed gene lov-1 (lovRT497F and lovRT993R), a positive control for male mRNA. Primer sequences for all experiments are provided in Supplementary Information.

RNAi construct and assay

A fragment of the plg-1 locus was amplified using primers e2.2F1 and plgR12BglII and the product cut with BglII and cloned into the RNAi feeding vector L4440 (Andy Fire Vector Kit, Addgene, Cambridge, MA). The insert is approximately 3 kb and contains almost the entirety of the large repetitive last exon of plg-1. The plasmid was then transformed into HT115(DE3) RNAse III-deficient bacteria for RNAi feeding experiments, which were conducted using the protocol described by Kamath and Ahringer31. Insertless vector L4440 in HT115 cells served as a negative control. Individual virgin N2 hermaphrodites were combined overnight with 3–5 CB4856 virgin males that were either exposed to bacteria containing insertless vector (these males should generate normal plugs), or exposed to bacteria producing plg-1 dsRNA (these males should generate fewer/smaller plugs). The hermaphrodites were scored for plugs the next morning, using an arbitrary plug size rating scale (0: none, 1: small, 2: medium, 3: large), and the observer was blind to treatment. CB4856 exhibits germline RNAi resistance, but it retains sensitivity to systemic RNAi in somatic tissues32, permitting us to use RNAi to assay gene function in the somatic gonad.

Transgenic complementation construct and transformation

We constructed the plg-1 rescue transgene pMR5.3 by sequentially cloning the 5′ and 3′ fragments. We cloned the 5′ part of the gene, amplified from CB4856 with primers plgF3 and plgR1 and cut with ApaI and BglII, into pPD34_110 (Andy Fire Vector Kit, Addgene), cut with the same enzymes. The 5′ insert is 1469 bp and includes 686 bp upstream of the 5′ end of the transcript as determined by RACE. We cloned the 3′ part of the gene, amplified with plgF1 and plgR1 and cut with XbaI and BglII, separately into pPD34_110 cut with those enzymes. This 3′ fragment from CB4856 is ~3.9 kb, versus 2.9 KB in the WormBase N2 genome sequence. We next excised the ApaI-BglII fragment from the first vector and ligated it into the second. We sequenced the resulting construct to the extent possible given the repetitive nature of the last plg-1 exon and found no differences from the sequence derived from CB4856 genomic DNA.

We introduced pMR5.3 and an unc-119 rescue plasmid, pMM016B, into strain QX1015, which carries unc-119(ed3) in a CB4856 background. Because unc-119 maps close to plg-1, QX1015 retains the N2 allele of plg-1 and other genes in this region, and N2/QX1015 F1 males fail to plug. We cobombarded 3μg of each plasmid into QX1015 using a BioRad gene gun. We recovered several independent transformed lines and selected one with high transmission of the extrachromosomal array, as judged by the scarcity of Unc worms in its progeny. This transformed line was assayed quantitatively for plug formation. Males of three genotypes were compared: CB4856, which is a natural plugging strain; F1 from a cross of N2 to QX1015 (plg-1(N2) unc-119(ed3) III, CB4856 background) carrying the extrachromosomal array [plg-1(CB4856) unc-119(+)], which are expected to produce plugs; and F1 from a cross of N2 to QX1015 without the array, which are expected to be unable to produce plugs. We combined 3–5 males overnight with a single N2 virgin hermaphrodite, and plug size was assessed the next morning in arbitrary units (0: none, 1: small, 2: medium, 3: large), with the observer blind to male genotype.

Promoter::GFP construct and transformation

We cloned a CB4856 PCR fragment encompassing the 311 bp 5′ of the start of translation, amplified using primers NotIplgprom471F and NplgproR1, into vector pSM_GFP using a NotI cut site. The plasmid, pMR18K, was introduced into QX1015 by biolistic transformation. We recovered a line with an integrated transgene, assigned allele name qqIs1. The transgene maps near the center of Chr III, near plg-1 and unc-119, based on linkage mapping with strains MT3751 (dpy-5 I; rol-6 II; unc-32 III), EG1020 (bli-6 IV; dpy-11 V; lon-2 X), and CB1562 (vab-7 III). We recovered 0/13 Vab GFP worms and 0/50 Unc-32 GFP worms. Several additional lines, in which the transgene is not integrated into the genome, show qualitatively identical GFP expression. Nematodes expressing GFP were imaged with a Zeiss 510 META confocal microscope using a Plan-NEOFLUAR 40X/1.3 oil objective. Nematodes were anesthetized in 10mM Sodium Azide and mounted on 5% agar pads for examination. Unmated males (e.g., Figure 3d–e) were collected as L4 larvae and held overnight without access to hermaphrodites. To obtain males soon after mating (e.g., Figure 3f), individual virgin males were placed in the presence of several virgin hermaphrodites and watched until matings were completed, then examined under the microscope immediately afterwards. We corroborated the relationship between mating and vacuole morphology by examining qqIs1; mab-23 homozygotes, which are incapable of mating. These males exhibited the large vacuole morphology.

Lectin Staining

Plugged and unplugged hermaphrodites were fixed in 4% paraformaldehyde in PBS. Fixed hermaphrodites were stained in 1:100 dilutions of the two molecules below (Invitrogen Molecular Probes), in PBS + 0.1% TritonX, then washed several times in PBS + 0.1% TritonX, and then several times in PBS.

  1. Lectin PNA from Arachis hypogaea (peanut), Alexa Fluor® 594 conjugate (Catalog Number - L32459)

Worms were mounted on 5% agar pads and examined under the confocal microscope. Strong lectin signal was observed; negligible signal was observed from the donkey anti-mouse antibody, suggesting that the plug was not simply absorbing probe.

Supplementary Material

Movie S1

Supplementary Information is linked to the online version of the paper at www.nature.com/nature.

Supplementary

Table S1

Acknowledgments

We thank the Caenorhabditis Genetics Center, M. Ailion, E. Dolgin, A. Barriere, and M.-A. Felix for strains. Work at Bowdoin College was supported by college funds, by the National Institutes of Health (grant P20 RR-016463 from the INBRE Program of the National Center for Research Resources), National Science Foundation (grant 0110994), and by an award to Bowdoin College by the Howard Hughes Medical Institute under the Undergraduate Science Education Program. Work at Princeton University was supported by grants from the National Institutes of Health (R01 HG004321 to LK and P50 GM071508 to the Lewis-Sigler Institute), a James S. McDonnell Foundation Centennial Fellowship (LK), and a Jane Coffin Childs Fellowship (MVR). We thank S. Civillico for assistance in the lab, F. Hagen for experimental advice, and E. Anderson, H. Coller, L. Gordon, and H. Seidel for comments on the manuscript.

References

1. Hodgkin J, Doniach T. Natural variation and copulatory plug formation in Caenorhabditis elegans. Genetics. 1997;146:149–64. [PMC free article] [PubMed]
2. Barker DM. Copulatory plugs and paternity assurance in the nematode Caenorhabditis elegans. Anim Behav. 1994;48:147–156.
3. Kiontke K, et al. Caenorhabditis phylogeny predicts convergence of hermaphroditism and extensive intron loss. Proc Natl Acad Sci U S A. 2004;101:9003–8. [PMC free article] [PubMed]
4. Cho S, Jin SW, Cohen A, Ellis RE. A phylogeny of Caenorhabditis reveals frequent loss of introns during nematode evolution. Genome Res. 2004;14:1207–20. [PMC free article] [PubMed]
5. Barriere A, Felix MA. High local genetic diversity and low outcrossing rate in Caenorhabditis elegans natural populations. Curr Biol. 2005;15:1176–84. [PubMed]
6. Barriere A, Felix MA. Temporal dynamics and linkage disequilibrium in natural Caenorhabditis elegans populations. Genetics. 2007;176:999–1011. [PMC free article] [PubMed]
7. Sivasundar A, Hey J. Sampling from natural populations with RNAi reveals high outcrossing and population structure in Caenorhabditis elegans. Curr Biol. 2005;15:1598–602. [PubMed]
8. Barriere A, Felix MA. Natural variation and population genetics of Caenorhabditis elegans. WormBook. 2005. http://www.wormbook.org. [PubMed] [Cross Ref]
9. Desseyn JL, Aubert JP, Porchet N, Laine A. Evolution of the large secreted gel-forming mucins. Mol Biol Evol. 2000;17:1175–84. [PubMed]
10. Perez-Vilar J, Hill RL. The structure and assembly of secreted mucins. J Biol Chem. 1999;274:31751–4. [PubMed]
11. Lang T, Alexandersson M, Hansson GC, Samuelsson T. Bioinformatic identification of polymerizing and transmembrane mucins in the puffer fish Fugu rubripes. Glycobiology. 2004;14:521–7. [PubMed]
12. Julenius K, Molgaard A, Gupta R, Brunak S. Prediction, conservation analysis, and structural characterization of mammalian mucin-type O-glycosylation sites. Glycobiology. 2005;15:153–64. [PubMed]
13. Bendtsen JD, Nielsen H, von Heijne G, Brunak S. Improved prediction of signal peptides: SignalP 3.0. J Mol Biol. 2004;340:783–95. [PubMed]
14. Dover G. Molecular drive: a cohesive mode of species evolution. Nature. 1982;299:111–7. [PubMed]
15. Lotan R, Skutelsky E, Danon D, Sharon N. The purification, composition, and specificity of the anti-T lectin from peanut (Arachis hypogaea) J Biol Chem. 1975;250:8518–23. [PubMed]
16. Carlson DM. Structures and immunochemical properties of oligosaccharides isolated from pig submaxillary mucins. J Biol Chem. 1968;243:616–26. [PubMed]
17. Lints R, Hall DH. Handbook of C. elegans male anatomy. WormAtlas. 2005. http://www.wormatlas.org/handbook/contents.htm.
18. Ganko EW, Fielman KT, McDonald JF. Evolutionary history of Cer elements and their impact on the C. elegans genome. Genome Res. 2001;11:2066–74. [PMC free article] [PubMed]
19. Britten RJ. Active gypsy/Ty3 retrotransposons or retroviruses in Caenorhabditis elegans. Proc Natl Acad Sci U S A. 1995;92:599–601. [PMC free article] [PubMed]
20. Barr MM, Sternberg PW. A polycystic kidney-disease gene homologue required for male mating behaviour in C. elegans. Nature. 1999;401:386–9. [PubMed]
21. Peden EM, Barr MM. The KLP-6 kinesin is required for male mating behaviors and polycystin localization in Caenorhabditis elegans. Curr Biol. 2005;15:394–404. [PubMed]
22. Barr MM, Garcia LR. Male mating behavior. WormBook. 2006. http://www.wormbook.org. [PubMed] [Cross Ref]
23. Fisher RA. The genetical theory of natural selection. Oxford University Press; Oxford: 1930.
24. Stern DL. Evolutionary developmental biology and the problem of variation. Evolution. 2000;54:1079–91. [PubMed]
25. Lints R, Emmons SW. Regulation of sex-specific differentiation and mating behavior in C. elegans by a new member of the DM domain transcription factor family. Genes Dev. 2002;16:2390–402. [PMC free article] [PubMed]
26. Chasnov JR, Chow KL. Why are there males in the hermaphroditic species Caenorhabditis elegans? Genetics. 2002;160:983–94. [PMC free article] [PubMed]
27. Cutter AD, Ward S. Sexual and temporal dynamics of molecular evolution in C. elegans development. Mol Biol Evol. 2005;22:178–88. [PubMed]
28. Stewart AD, Phillips PC. Selection and maintenance of androdioecy in Caenorhabditis elegans. Genetics. 2002;160:975–82. [PMC free article] [PubMed]
29. Brudno M, et al. LAGAN and Multi-LAGAN: efficient tools for large-scale multiple alignment of genomic DNA. Genome Res. 2003;13:721–31. [PMC free article] [PubMed]
30. Sulston JE, Hodgkin J. The nematode Caenorhabditis elegans. Cold Spring Harbor Press; Cold Spring Harbor: 1988.
31. Kamath RS, Ahringer J. Genome-wide RNAi screening in Caenorhabditis elegans. Methods. 2003;30:313–21. [PubMed]
32. Tijsterman M, Okihara KL, Thijssen K, Plasterk RH. PPW-1, a PAZ/PIWI protein required for efficient germline RNAi, is defective in a natural isolate of C. elegans. Curr Biol. 2002;12:1535–40. [PubMed]
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...