We previously demonstrated that sperm heads from amphibians (Xenopus and Rana) and zebrafish (Danio) could form giant lampbrush chromosomes when injected into the nucleus of amphibian oocytes. However, similar experiments with mammalian sperm heads were unsuccessful. Here, we describe a slightly modified procedure and demonstrate that human sperm heads can form giant lampbrush chromosomes when injected into the oocyte nucleus of the frog Xenopus laevis or the newt Notophthalmus viridescens. Human and other mammalian chromosomes do not form recognizable lampbrush chromosomes in their own oocytes or in any somatic cells. These experiments thus demonstrate that the lampbrush condition is an inducible state and that the amphibian oocyte nucleus contains all factors required to remodel the inactive chromatin of a mammalian sperm into a transcriptionally active state. They also demonstrate that absence of lampbrush chromosomes from human oocytes must relate to specific features of mammalian oogenesis, not to permanent genetic or epigenetic changes in the chromatin.
Keywordschromatin oocyte reprogramming RNA polymerase II sperm transcription Xenopus
- Pol II
RNA polymerase 2
- Pol III
RNA polymerase 3
Lampbrush chromosomes (LBCs) are giant meiotic chromosomes first described more than 100 years ago from the oocyte nucleus or germinal vesicle (GV) of the axolotl (Flemming 1882) and a shark (Rückert 1892). LBCs occur in many species, including vertebrates, invertebrates, and even the single-celled alga Acetabularia, but they are best characterized from amphibian and avian oocytes (Callan and Lloyd 1960; Morgan 2002; Gall et al. 2004; Gaginskaya et al. 2009). Each homologue has an axis of chromomeres corresponding to transcriptionally inactive chromatin, from which transcriptionally active pairs of loops extend laterally. A coating of nascent ribonucleoprotein (RNP) fibrils makes these lateral loops visible by conventional light microscopy. Because of their enormous size, LBCs provide an ideal system in which to study transcription, RNA processing, and other general features of chromosome organization. The overall organization and significance of LBCs were summarized in the classic monograph by Mick Callan (1986) (see also www.projects.exeter.ac.uk/lampbrush).
In a previous study, we demonstrated that demembranated sperm heads from the frogs Xenopus and Rana and the zebrafish Danio can form LBCs when injected into the GV of Xenopus laevis (Gall and Murphy 1998). Although they are unreplicated single chromatids, sperm LBCs are similar to endogenous lampbrush bivalents in morphology and immunofluorescent staining properties. The induction of sperm LBCs in amphibian GVs provides a useful system for identifying cis- and trans-acting factors required for converting condensed chromatin into a transcriptionally active form. In our earlier experiments, we failed to induce LBCs from mammalian sperm heads. It was unclear whether the failure was due to technical issues or to more fundamental differences between amphibian and mammalian chromatin. Because mammalian chromosomes do not form recognizable LBCs during meiosis or in any somatic cells, one could postulate that mammalian chromatin is unable to assume the LBC condition. Here, we describe new experiments under slightly different conditions in which human sperm heads give rise to transcriptionally active LBCs when injected into the GV of the frog X. laevis or the newt Notophthalmus viridescens. These experiments demonstrate that the amphibian GV contains all factors required to reprogram inactive mammalian chromatin into a transcriptionally active state. Thus, the absence of LBCs from mammalian oocytes must relate to specific aspects of mammalian oogenesis and not to permanent genetic or epigenetic features of mammalian chromatin.
Materials and methods
Oocytes and LBC spreads
Adult frogs X. laevis were purchased from Xenopus 1 (Dexter, MI) and adult newts N. viridescens from the Sullivan Company (Nashville, TN). Oocytes were held at 16–18 °C or 24 °C in a small Petri dish of OR2 saline (Wallace et al. 1973). Chromosome spreads were prepared from individual GVs as described previously (Gall and Wu 2010).
Sperm heads of X. laevis were prepared from testes as described earlier (Newmeyer and Wilson 1991). A sample of normal human sperm was obtained from the Johns Hopkins University Medical School. The research was deemed by the Johns Hopkins Institutional Review Board to qualify for exemption under category (4) of 45 CFR 46.101(b). The sperm tails were removed by sonication, and the heads were concentrated by centrifugation. The technique for injection of human sperm heads into the GV was basically as described previously for experiments with Xenopus, Rana, and Danio (Gall and Murphy 1998) with four relatively minor changes. Although the effect of any one of these changes is difficult to assess, the new protocol now gives a higher survival rate and allows us to hold oocytes for longer periods before observing the GV contents. A major difference between Xenopus and human sperm heads is the longer time needed for the human sperm heads to generate recognizable LBCs.
First, in our previous study we injected sperm heads that had been demembranated with lysolecithin, as originally described by Gurdon (1976). In the experiments described here, we found that intact sperm heads started swelling and eventually resolved into individual LBCs at a pace comparable to that of demembranated sperm heads. For this reason, we omitted the lysolecithin step.
Second, we did not defolliculate the oocytes with collagenase before injection. Defolliculated oocytes are softer and easier to penetrate with the injection needle, but are more fragile and prone to contamination. By using sharp needles with smaller tips, we managed to inject oocytes successfully without defolliculation.
Third, we did not centrifuge the oocytes before injection. The major purpose of centrifugation is to bring the GV to the surface, where its position can be detected as a depigmented area. With practice, one can inject sperm heads into the GV even when it lies deeper within the oocyte.
Finally, in earlier experiments we found that the number of sperm heads actually injected decreased dramatically during the course of injecting multiple oocytes. We reasoned that this might be due to adherence of sperm heads to the inside wall of the needle. To prevent such sticking, we added 2–5 % polyvinylpyrrolidone (PVP) to the sperm suspension. Subsequently, we found that the number of sperm actually injected remained relatively constant during the course of an experiment. PVP has been used routinely in intracytoplasmic sperm injections involving mammalian sperm and oocytes.
After improving the injection and oocyte handling protocol, we were able to keep injected oocytes alive for several days. The longer incubation time is critical because human sperm heads expand more slowly than Xenopus sperm heads. In our previous experiments, most of the injected oocytes were beginning to degenerate before the human sperm heads had expanded.
Immunofluorescence staining and microscopy
GV spreads were stained with antibodies as described previously (Gall and Murphy 1998) with the following modifications. Samples were blocked with 10 % horse serum for 30 min before incubation in primary antibody for 4–12 h at room temperature. Primary antibodies used in this study include mouse mAb H14 against RNA polymerase 2 (pol II) (Bregman et al. 1995); mouse mAb Y12 against the Sm “epitope” (symmetric dimethylarginine) (Lerner et al. 1981; Brahms et al. 2000); and mouse mAb H1 against Xenopus coilin (Tuma et al. 1993). Secondary antibodies were Alexa 488- or Alexa 594-conjugated goat anti-mouse IgG or IgM (Molecular Probes, Eugene, OR). They were used together with 0.1 μg/ml 4′,6 diamidino-2-phenylindole (DAPI) for 4–12 h at room temperature. Slides were mounted in 50 % glycerol. Specimens were examined with a Zeiss 63X 1.25 N.A. planapo lens on the Zeiss Axioplan fluorescence microscope. Images were captured with a Micromax charge-coupled device camera (Princeton Instruments, Trenton, NJ) using the IPLab (3.5.5) image acquisition and analysis program (Scanalytics, Fairfax, VA).
Injection of human sperm heads into the Xenopus GV
Temperature accelerates the time for formation of LBCs
Injection of human sperm heads into the newt GV
Association of nuclear bodies with human LBCs
It is also significant that the bodies associated with the human LBCs stain with anti-coilin mAb H1. This antibody is highly species-specific for Xenopus, not staining coilin in either newt or human cells. Thus, it is clear that the coilin in the nuclear bodies associated with the human LBCs is Xenopus coilin. In our earlier injection experiments, we showed that the loops on Xenopus LBCs derived by injection of Xenopus sperm heads into newt GVs stained strongly with an antibody against a newt-specific protein. Both examples demonstrate that endogenous proteins are used for assembly of the induced LBCs and associated bodies.
The major finding of this study is that condensed human chromatin from sperm heads is able to form typical LBCs when placed in the environment of the amphibian oocyte nucleus. Although a formal definition of a LBC is difficult to make, for purposes of the discussion here we mean a giant chromosome with transcriptionally active lateral loops visible by conventional light microscopy. Our earlier experiments involved a detailed analysis of the LBCs formed when X. laevis sperm heads were injected into the GV of the same species. We showed that the induced LBCs were identical to endogenous LBCs in all essential respects, with the exception that they consisted of single, unpaired chromatids rather than meiotic bivalents. We also showed that sperm heads of the frog Rana pipiens and the zebrafish Danio rerio transformed into recognizable univalent LBCs when injected into the GV of X. laevis. However, experiments with cricket (Acheta domesticus), mouse (Mus musculus), and human sperm heads were unsuccessful, leaving unanswered the question whether the source of the inactive chromatin is important. Our original experiments involved three species that normally have LBCs in their oocyte nuclei, whereas human chromosomes do not go through a LBC stage during oocyte development (or at any other time). Our experiments demonstrate that the absence of LBCs from mammalian oocytes must be due to specific features of mammalian oogenesis and not to permanent genetic or epigenetic features of mammalian chromatin.
Now that we have obtained positive results with human sperm heads, we feel confident in predicting that chromatin from essentially any sperm or germ line source can be converted to the LBC state, so long as it can be made accessible to factors in the GV. Whether chromatin from fully differentiated somatic cells can be similarly converted to the LBC state remains to be demonstrated.
The importance of factors in the GV, as opposed to the source of the chromatin, is underscored by comparing injections into frog and newt GVs. In our earlier experiments, we showed that induced Xenopus LBCs were larger and had more prominent lateral loops when formed in the newt GV (Notophthalmus) than in the endogenous frog GV (Xenopus). The same is true in our current experiments with human LBCs. A comparison of the LBCs in Figs. 1 and 4 with those in Fig. 3 shows a dramatic difference. It is well known that LBCs of the newt are much larger than those of Xenopus, and it has frequently been assumed that this difference is related to the large difference in total genomic DNA of the two species (3 versus 35 pg in the haploid genome). Our injection experiments raise the interesting possibility that at least some of the differences in LBC size are related to specific factors in the GV, not simply to differences in genomic DNA content.
At present, we have very little information concerning the endogenous factors in the GV that are involved. We know that pol II goes onto the chromatin at an early stage, but we assume that the chromatin itself must be modified before transcription can begin. The GV and its giant chromosomes provide a uniquely favorable system in which to study both the changes that occur during reprogramming of the chromatin and the factors in the nucleoplasm that are responsible for the reprogramming.
The ideas presented here owe much to discussions with Svetlana Deryusheva, Garry Morgan, Zehra Nizami, and Jun Wei Pek. This work was supported by research grant R01 GM33397 from the National Institute of General Medical Sciences of the National Institutes of Health. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. JGG is American Cancer Society Professor of Developmental Genetics.
- Brahms H, Raymackers J, Union A, de Keyser F, Meheus L, Lührmann R (2000) The C-terminal RG dipeptide repeats of the spliceosomal Sm proteins D1 and D3 contain symmetrical dimethylarginines, which form a major B-cell epitope for anti-Sm autoantibodies. J Biol Chem 275:17122–17129PubMedCrossRefGoogle Scholar
- Flemming W (1882) Zellsubstanz, Kern und Zelltheilung. F. C. W. Vogel, LeipzigGoogle Scholar
- Gall JG, Diaz MO, Stephenson EC, Mahon KA (1983) The transcription unit of lampbrush chromosomes. In: Subtelny S, Kafatos F (eds) Gene structure regulation in development. Alan R. Liss, New York, pp 137–146Google Scholar
- Nizami ZF, Gall JG (2012) Pearls are novel Cajal body-like structures in the Xenopus germinal vesicle that are dependent on RNA pol III transcription. Chromosome Res. doi:10.1007/s10577-012-9320-1
- Rückert J (1892) Zur Entwickelungsgeschichte des Ovarialeies bei Selachiern. Anat Anz 7:107–158Google Scholar