"Genome-wide non-mendelian inheritance of extra-genomic information in Arabidopsis".

Susan J. Lolle*, Jennifer L. Victor, Jessica M. Young,  and Robert E. Pruitt*

Department of Botany and Plant Pathology, Purdue University, 915 W. State Street, West Lafayette, Indiana 47907-2054, USA
* These authors contributed equally to this work

Correspondence and requests for materials should be addressed to R.E.P. ( pruittr@purdue.edu ).

A fundamental tenet of classical mendelian genetics is that allelic information is stably inherited from one generation to the next, resulting in predictable segregation patterns of differing alleles [1]. Although several exceptions to this principle are known, all represent specialized cases that are mechanistically restricted to either a limited set of specific genes (for example mating type conversion in yeast [2]) or specific types of alleles (for example alleles containing transposons [3] or repeated sequences [4]). Here we show that Arabidopsis plants homozygous for recessive mutant alleles of the organ fusion gene HOTHEAD[5] (HTH) can inherit allele-specific DNA sequence information that was not present in the chromosomal genome of their parents but was present in previous generations. This previously undescribed process is shown to occur at all DNA sequence polymorphisms examined and therefore seems to be a general mechanism for extra-genomic inheritance of DNA sequence information. We postulate that these genetic restoration events are the result of a template-directed process that makes use of an ancestral RNA-sequence cache.

We have recovered 11 point mutations at the hth locus [6] that share the unusual property that they segregate phenotypically wild-type plants at a high frequency when homozygous mutant plants are allowed to self-fertilize. The frequency with which these 'revertant' plants are recovered varies, but is generally in the range of 10^-1 to 10^-2 revertants per chromosome per generation (Table 1). This is in stark contrast to most point mutations, which are completely stable (for example erecta (er) in Table 1). Because of the high frequency with which hth phenotypic revertants were observed, we suspected that they might reflect either incomplete penetrance of the mutant phenotype or an epigenetic change that masked the mutant genotype. To determine the genotype of the phenotypic revertants we designed allele-specific polymerase chain reaction (PCR)-based markers. These molecular markers show clearly that the phenotypically wild-type plants observed are heterozygous for the parental hth allele (Fig. 1a, Table 1), indicating that the nucleotide sequence of the hth gene had been altered in about 10% of the progeny plants. This was confirmed by allowing the revertant plants to self-fertilize and examining the segregation of the hth phenotype in the next generation. In every case the mutant phenotype was observed in about 25% of the progeny plants (data not shown).

Table 1:

a, PCR-based assay demonstrating genetic reversion of hth-8. DNA was extracted from mutant and revertant progeny that were derived from the self-fertilization of a homozygous hth-8 parent, amplified with gene-specific primers and digested to reveal the presence or absence of a polymorphic restriction site. Controls in the left three lanes show digest patterns derived from wild-type (HTH/HTH), heterozygous (HTH/hth) and homozygous (hth/hth) mutant plants. Photos above mutant and revertant lanes show the hth/hth mutant phenotype and the revertant phenotype.

b, Genome blot showing that HTH is present in a single copy in Col (C), Ler (L) and Ws (W). Genomic DNA was digested as indicated and probed with a labelled fragment of Col DNA corresponding to the HTH coding sequence. Expected fragment sizes in base pairs are as follows: for HindIII: Col 1,285, 1,253, 664; Ler 1,275, 1,253, 664; Ws 1,253, 1,204, 666; for ScaI: Col 2,537, 2,187; Ler 2,527, 2,187; Ws 2,466, 2,188; for XhoI: Col 3,928, 3,016; Ler 3,918, 3,016, Ws 3,669. Differences in hybridization intensity occur because portions of the Ws sequence have diverged significantly from the Col sequence.

Two trivial explanations could account for the observed revertant progeny: contamination of our stocks with wild-type seed or outcrossing of the hth mutant plants with wild-type pollen. Seed contamination was an unlikely explanation because the apparent revertants are all heterozygous for the specific hth allele carried by their parents. To rule out all possibility of seed contamination, individual embryos were dissected out of fruits developing on selfed plants that were genotypically confirmed to be hth/hth by using allele-specific PCR-based markers. The genotypes of these embryos with respect to HTH were then determined with the same markers. The results in Table 2 show that we can detect frequent HTH/hth embryos developing on hth/hth parent plants, eliminating the possibility that the observed high frequency of reversion is due to seed contamination. In fact, in this experiment we also detected rare double revertant HTH/HTH embryos, which must have inherited one of their two wild-type HTH genes from the maternal parent and therefore could not have been the result of outcrossing. The frequency of apparent reversion events observed in embryos is somewhat higher than that observed phenotypically. We believe this probably reflects somatic reversion events occurring in embryonic tissues that do not contribute to the adult plant.

Table 2:

To demonstrate directly that hth/hth mutant plants can be a source of wild-type HTH alleles, we tested whether it was possible to transmit an HTH allele through pollen derived from an hth/hth plant. Plants homozygous for hth-4 in the Landsberg erecta (Ler) background were reciprocally crossed with wild-type Columbia (Col) plants and the embryos arising from the crosses were genotyped with respect to HTH. Most embryos were heterozygous as expected, but in the crosses where the male parent was homozygous for hth-4, 8 of 164 embryos were observed that were of the genotype HTH/HTH (reversion frequency 4.9 times 10^-2). Because the Col and Ler accessions differ at numerous DNA sequence polymorphisms, we used accession-specific DNA markers to verify that the HTH/HTH embryos were not due to self-pollination of the Col parent. No HTH/HTH embryos were detected when the female parent was homozygous for hth-4 (0 of 230 embryos). These data indicate that hth/hth plants can be a source of pollen bearing an HTH allele and furthermore that there is a strong bias for genetic changes to occur in the male reproductive system. Taken together, the results show that the wild-type HTH alleles observed in progeny of selfed hth/hth mutant plants are not the result of seed contamination or cross-pollination but represent true genetic events manifested at the DNA sequence level.

We considered several conventional explanations for the genetic instability observed at hth. DNA sequence analysis of the HTH gene rules out the involvement of transposons and repeated sequences as possible explanations. Two additional explanations that we considered were a high rate of random mutation in this particular region of the genome and correction of the gene through a gene conversion mechanism.

If reversion of hth to HTH were due to a high rate of random mutation at this locus, we would expect to find additional silent sequence changes in the reverted HTH allele. Given the frequency with which the mutant nucleotide is observed to revert to wild-type and the 1,226 possible silent nucleotide substitutions in the HTH coding sequence, we would expect to see between 49 and 100 other changes on average in the coding sequence of the reverted alleles, assuming that all types of nucleotide substitutions are equally likely. To test this we determined the complete DNA sequence of the coding region of the HTH gene after three independent reversion events in each of three independently derived mutant hth alleles (hth-4, hth-8 and hth-10). In every case the sequence of the reverted HTH allele matched the Ler wild-type sequence exactly. Given the very high frequency with which the mutant nucleotide reverts to the wild-type sequence, this result rules out a mechanism dependent on random sequence changes and strongly indicates a mechanism that is template-directed.

In considering the possibility of gene conversion we needed to exclude two types of source template. First, it was possible that the Arabidopsis Ler and Wassilewskija (Ws) accessions contained a second copy of the HTH gene that was not present in the Col genome sequence. This is unlikely because we have designed many PCR-based genetic markers that allow genotyping of mutant hth alleles and none revealed a cryptic wild-type copy of the gene in homozygous hth plants. To examine this possibility further, we performed a DNA blotting experiment with genomic Col, Ler and Ws DNA. We were able to detect all of the expected HTH fragments in this experiment, and no additional hybridizing sequences were identified (Fig. 1b). Taken together, the PCR and genomic hybridization results effectively rule out the existence of a second cryptic copy of the HTH gene.

The second class of template we considered was the family of HOTHEAD-LIKE (HTL) genes [6]. For a variety of reasons this possibility could also be discounted. For hth-4, none of the HTL sequences contained the appropriate nucleotide to correct the mutant sequence to wild-type (Fig. 2). For hth-8 and hth-10, some of the HTL sequences did contain the relevant nucleotide to correct the mutant sequence, but if any of these related sequences had been used as a corrective template this would have resulted in the incorporation of other nearby DNA sequence changes, unless the conversion tract were limited to just a few nucleotides (Fig. 2). Because no other DNA sequence changes were observed, it seems unlikely that conventional gene conversion is the explanation for the high level of genetic change seen at the HTH locus.

Figure 2: DNA sequences of HTH and HTL genes around the sites of mutation in hth-4, hth-8 and hth-10.

Figure 2 DNA sequences of HTH and HTL genes around the sites of mutation in hth-4, hth-8 and hth-10.

The location of the mutant nucleotide is boxed, and the wild-type nucleotide at this position is shown in blue; differences from the HTH sequence are shown in red lower-case letters. For hth-4 none of the related sequences contain the wild-type nucleotide at the site of the mutation, for hth-8 all of the other sequences contain the appropriate wild-type nucleotide, and for hth-10 some of the other sequences contain the relevant wild-type nucleotide. In all cases there are sequence differences between the HTL sequences and HTH within a few nucleotides of the site of the mutation.

In the experiments described above we examined the process of DNA sequence change at the hth locus in an hth mutant background. To determine whether DNA sequence changes also occur elsewhere in the genome, we crossed hth-4, hth-8 and hth-10 mutant plants (in the Ler genetic background) to wild-type Col plants. The F1 hybrids were allowed to self-fertilize and individual F2 plants were genotyped with the use of PCR-based markers to identify those that were homozygous for the mutant or wild-type allele of HTH. These F2 individuals were then genotyped for several molecular markers that are polymorphic between Col and Ler, to identify individuals that were homozygous for one parental allele or the other. Randomly chosen F3 progeny from these plants were in turn examined for the stable inheritance of those DNA sequence polymorphisms. In progeny of hth/hth F2 plants we observed a high rate of sequence change in both directions (Col to Ler, and Ler to Col), indicating that the mechanism of genetic instability observed at hth seems to apply globally to other sequences in the genome (Table 3). Although the polymorphisms examined are all within genes, their locations are in exons, introns and the 3' untranslated region of those genes (Table 3), indicating that presence in the mature mRNA is not required for extra-genomic transmission of sequence information. We also detected phenotypic reversion of the er mutation in the progeny of hth/hth er/er F2 plants at similar frequencies, demonstrating that the absence of HTH leads to the destabilization of er, which is in contrast to the stability seen in an HTH/HTH background (Table 1). Similarly, inheritance of molecular markers in the progeny of the HTH/HTH F2 plants was also completely stable, indicating that the instability observed is dependent on the absence of HTH gene function.

Table 3:

Mutation of the Arabidopsis HTH gene reveals an unusual pattern of genetic transmission in which progeny plants inherit, at a relatively high frequency, DNA sequences different from those carried by their parents. Although this was initially observed as a result of the instability of nucleotide substitution mutations in the hth gene itself, we have subsequently shown that loss of HTH gene function also results in the genetic instability of DNA sequences located throughout the Arabidopsis genome. However, the instability observed is not random but seems to be confined to the restoration of sequence information in progeny plants that, although absent from the parent genome, was present in the genome of an earlier ancestor. Thus, we believe that mutant hth alleles are able to revert to wild type because recent ancestors of the hth/hth mutant plants had a wild-type copy of the HTH gene and, similarly, that we can identify sequence polymorphism changes in plants whose parents were homozygous for one form of the polymorphism because they had grandparents that were heterozygous for the same polymorphism. Sequencing of multiple independently reverted hth alleles revealed no DNA sequence changes other than the restoration of the mutant nucleotide to the wild type, also indicating that the process of sequence change is non-random.

Similar patterns of inheritance have been observed previously under various exceptional circumstances. It has been reported that molecular markers absent from several generations of a population of synthetic polyploid plants can reappear subsequently in the population [7]. It has also been reported that random amplified polymorphic DNA (RAPD) markers can in rare instances be transmitted from stock to scion after grafting [8]. In addition to reports of aberrant inheritance of molecular markers, there are also numerous reports of 'replicating instabilities', which seem to propagate both wild-type and mutant alleles in a single copy of a gene [9]. All of these occurred at much lower frequencies than we have observed in the hth mutant background, but could be explained by a common mechanism involving the non-mendelian inheritance of sequence information maintained outside the conventional DNA genome.

On the basis of the high degree of specificity seen in the observed DNA sequence changes, we propose that this is a template-directed process. Because we have been unable to detect a DNA template by either genomic DNA blotting or PCR, a logical alternative is that the template is instead provided by an RNA molecule. We propose a model in which a type of stable RNA, possibly in a double-stranded form, can be replicated and transmitted over multiple generations. This template must under some circumstances be capable of modifying the DNA sequence of the nuclear genome to restore sequence information cached from previous generations. Given that this type of sequence modification occurs globally throughout the genome, we think it likely that the postulated genome-wide sequence cache also exists in wild-type plants and that the hth mutation simply increases the frequency with which DNA sequences are modified.

Although the model proposed above might represent an extraordinary view of inheritance, each of the processes required by the model has been demonstrated previously. Research performed over the past few years in post-transcriptional gene silencing or RNA interference has shown that silencing induced by double-stranded RNA can persist for several generations in Caenorhabditis elegans [10] and can spread through the organism in plants, fungi and animals [11, 12]. This strongly implies that some form of nucleotide sequence information must be stably replicated and transmitted outside the DNA genome, which is consistent with a requirement for an RNA-directed RNA polymerase in post-transcriptional gene silencing [13-15]. Replication of the sequence cache could allow it to persist for several generations; this is consistent with the fact that hth mutations retain their ability to revert even after multiple generations in the homozygous state (S.J.L. and R.E.P., unpublished observations). Experiments to detect changes in the reversion frequency after several generations of homozygous propagation are in progress. Both double-stranded RNA and microRNA have been implicated in the sequence-specific modification of the nuclear genome by DNA methylation [16-18] and it has been possible to perform site-directed mutagenesis by the introduction of chimaeric RNA–DNA oligonucleotides in both plant and animal cells [19-21]. We propose that a mechanism combining these individual steps could explain the exceptional inheritance we see in hth mutants as well as the unusual results observed by earlier workers.

It is tempting to speculate on the function of this unusual mechanism of inheritance in the wild-type organism. On the basis of our observation that progeny of hth mutant plants seem to inherit a wide variety of sequence information missing from the DNA genomes of their parents, one possibility is that this mechanism provides a cache of genetic variation beyond that carried in the chromosomes. If this information can then be copied back into the DNA genome it could allow self-fertilizing species such as Arabidopsis to avoid the negative consequences of inbreeding. Perhaps even more intriguing is the possibility that the inheritance of non-genomic templates occurs in wild-type and hth plants, but the rate at which those templates are used to modify genomic sequences is elevated in hth as a result of an indirect 'stress' put on the plant by the absence of the HTH gene product. Under these circumstances one could envisage a mechanism in which additional allelic information is maintained outside the normal genomic context but could be used under conditions that compromised the continued functioning of the organism. The extra-genomic information used by the organism would not be random in this case but would contain a library of potential allelic sequences that were previously carried in the DNA of its ancestors, all of which must have survived to reproduce. These sequences would therefore represent a collection of alleles already known to be functional under at least some circumstances, and the restoration of these sequences to the DNA genome might represent a mechanism for the organism to recover from a genotype that was poorly adapted to its present environment. Any sequence changes that resulted in the elimination of the 'stress' would also decrease the rate of sequence alteration back to its normal background level, effectively fixing a favourable genotype. This is at least superficially similar to some models proposed to account for 'adaptive' mutation in microorganisms [22-24] and it is possible that recent work describing increased rates of genomic change in response to environmental stress [25, 26] and pathogen attack [27] reflect the action of such a mechanism. It is also possible that some of the cryptic genetic variation uncovered through the loss of Hsp90 function [28, 29] is actually present in the postulated ancestral sequence cache and is restored to the DNA genome in response to a lower level of Hsp90 activity.

Whatever molecular mechanism leads to the unusual pattern of genetic inheritance we have observed in hth mutants, the genetic results presented here indicate the existence of an extra-genomic mechanism that leads to the high-frequency modification of DNA sequences in a template-directed manner. The structure of the templates together with the mechanisms by which they are produced, inherited and used remain questions for future research. It will be of great interest to see how widely this mechanism of inheritance is distributed throughout the tree of life.

Methods
DNA extraction, blotting, labelling and hybridization techniques were as described previously [5]. Genomic DNA sequences were determined by amplifying portions of the hth gene and directly sequencing the PCR product [30]. DNA samples for the PCR genotyping of embryos were obtained by dissecting individual embryos from siliques and surrounding maternal seed coat tissue, crushing the embryo with a plastic pestle in a Microfuge tube and resuspending the resulting homogenate in 20 µl of TE buffer (10 mM Tris-HCl pH 8.0, 1 mM EDTA). A 1 µl portion of this sample was used in a standard PCR reaction.

Received 8 December 2004;accepted 21 January 2005
------------------

References

1. Mendel, G. Versuche über Pflanzen Hybriden. Verhandl. Naturforsch. Ver. Brünn 4, 3–47 (1866)

2. Klar, A. J., Fogel, S. & Lusnak, K. Gene conversion of the mating-type locus in Saccharomyces cerevisiae. Genetics 92, 777–782 (1979)

3. McClintock, B. The origin and behavior of mutable loci in maize. Proc. Natl Acad. Sci. USA 36, 344–355 (1950)

4. Gondo, Y. et al. High-frequency genetic reversion mediated by a DNA duplication: the mouse pink-eyed unstable mutation. Proc. Natl Acad. Sci. USA 90, 297–301 (1993)

5. Lolle, S. J., Hsu, W. & Pruitt, R. E. Genetic analysis of organ fusion in Arabidopsis thaliana. Genetics 149, 607–619 (1998)

6. Krolikowski, K. A., Victor, J. L., Nussbaum Wagler, T., Lolle, S. J. & Pruitt, R. E. Isolation and characterization of the Arabidopsis organ fusion gene HOTHEAD. Plant J. 35, 501–511 (2003)

7. Song, K., Lu, P., Tang, K. & Osborn, T. C. Rapid genome change in synthetic polyploids of Brassica and its implications for polyploid evolution. Proc. Natl Acad. Sci. USA 92, 7719–7723 (1995)

8. Taller, J., Hirata, Y., Yagishita, N., Kita, M. & Ogata, S. Graft-induced genetic changes and the inheritance of several characteristics in pepper (Capsicum annuum L.). Theor. Appl. Genet. 97, 705–713 (1998)

9. Auerbach, C. & Kilbey, B. J. Mutation in eukaryotes. Annu. Rev. Genet. 5, 163–218 (1971)

10. Fire, A. et al. Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature 391, 806–811 (1998)

11. Cogoni, C. et al. Transgene silencing of the al-1 gene in vegetative cells of Neurospora is mediated by a cytoplasmic effector and does not depend on DNA-DNA interactions or DNA methylation. EMBO J. 15, 3153–3163 (1996)

12. Palauqui, J. C., Elmayan, T., Pollien, J. M. & Vaucheret, H. Systemic acquired silencing: transgene-specific post-transcriptional silencing is transmitted by grafting from silenced stocks to non-silenced scions. EMBO J. 16, 4738–4745 (1997)

13. Cogoni, C. & Macino, G. Gene silencing in Neurospora crassa requires a protein homologous to RNA-dependent RNA polymerase. Nature 399, 166–169 (1999)

14. Dalmay, T., Hamilton, A., Rudd, S., Angell, S. & Baulcombe, D. C. An RNA-dependent RNA polymerase gene in Arabidopsis is required for posttranscriptional gene silencing mediated by a transgene but not by a virus. Cell 101, 543–553 (2000)

15. Mourrain, P. et al. Arabidopsis SGS2 and SGS3 genes are required for posttranscriptional gene silencing and natural virus resistance. Cell 101, 533–542 (2000)

16. Wassenegger, M., Heimes, S., Riedel, L. & Sanger, H. L. RNA-directed de novo methylation of genomic sequences in plants. Cell 76, 567–576 (1994)

17. Mette, M. F., Aufsatz, W., van der Winden, J., Matzke, M. A. & Matzke, A. J. Transcriptional silencing and promoter methylation triggered by double-stranded RNA. EMBO J. 19, 5194–5201 (2000)

18. Bao, N., Lye, K.-W. & Barton, M. K. MicroRNA binding sites in Arabidopsis class III HD-ZIP mRNAs are required for methylation of the template chromosome. Dev. Cell 7, 653–662 (2004)

19. Cole-Strauss, A. et al. Correction of the mutation responsible for sickle cell anemia by an RNA-DNA oligonucleotide. Science 273, 1386–1389 (1996)

20. Beetham, P. R., Kipp, P. B., Sawycky, X. L., Arntzen, C. J. & May, G. D. A tool for functional plant genomics: chimeric RNA/DNA oligonucleotides cause in vivo gene-specific mutations. Proc. Natl Acad. Sci. USA 96, 8774–8778 (1999)

21. Zhu, T. et al. Targeted manipulation of maize genes in vivo using chimeric RNA/DNA oligonucleotides. Proc. Natl Acad. Sci. USA 96, 8768–8773 (1999)

22. Hall, B. G. Spontaneous point mutations that occur more often when advantageous than when neutral. Genetics 126, 5–16 (1990)

23. Foster, P. L. & Cairns, J. Mechanisms of directed mutation. Genetics 131, 783–789 (1992)

24. Galitski, T. & Roth, J. R. A search for a general phenomenon of adaptive mutability. Genetics 143, 645–659 (1996)

25. Ries, G. et al. Elevated UV-B radiation reduces genome stability in plants. Nature 406, 98–101 (2000)

26. Bjedov, I. et al. Stress-induced mutagenesis in bacteria. Science 300, 1404–1409 (2003)

27. Kovalchuk, I. et al. Pathogen-induced systemic plant signal triggers DNA rearrangements. Nature 423, 760–762 (2003)

28. Queitsch, C., Sangster, T. A. & Lindquist, S. Hsp90 as a capacitor of phenotypic variation. Nature 417, 618–624 (2002)

29. Rutherford, S. L. & Lindquist, S. Hsp90 as a capacitor for morphological evolution. Nature 396, 336–342 (1998)

30. Pruitt, R. E., Vielle-Calzada, J. P., Ploense, S. E., Grossniklaus, U. & Lolle, S. J. FIDDLEHEAD, a gene required to suppress epidermal cell interactions in Arabidopsis, encodes a putative lipid biosynthetic enzyme. Proc. Natl Acad. Sci. USA 97, 1311–1316 (2000)

Acknowledgements. We thank R. M. Lee and T. Nussbaum Wagler for technical assistance, and J. A. Banks, J. L. Bowman, U. Grossniklaus, R. A. Jorgensen, J. Ogas, D. R. Smyth, A. Steimer, V. Sundaresan and M. A. Webb for discussions. This work was supported by the National Science Foundation. Any opinion, findings, conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.

Competing interests statement. The authors declare that they have no competing financial interests.

1. Kiyosawa H, Mis N, Iwase S, Hayashizaki Y,  and Abe K, "Disclosing hidden transcripts: Mouse natural sense-antisense transcripts tend to be poly(A) negative and nuclear localized". Published online before print March 21, 2005 Genome Research, DOI: 10.1101/gr.3155905

2. Love R, "Distribution of Ribonucleic Acid in Tumor Cells during Mitosis", Nature vol. 180, no. 4598, pp. 1338-1339 (December 14, 1957).

3. De Carvalho S, "Effect of RNA from Normal Human Marrow on Leukaemic Marrow In-Vivo", Nature, vol. 197, no. 4872, pp. 1077-1080, (March 16, 1963).

Links to RNA and Biological Causality:

Links to Euchromatin Activator RNA Reviews:
Links to Euchromatin Activator RNA Research:
Links to Ultrastructural Probes of DNase I-Sensitive Sites:
Links to RNA as a Therapeutic Agent:
Links to Hodgkin Lymphoma Immuno-Pathology:
Links to Activated T-Lymphocyte Immunotherapy:
Links to Medical Systems Biology:
Links to Selective Gene Transcription:
Links to RNA-Induced Epigenetics:
Links to RNA-Induced Embryogenesis:
Links to RNA and Biological Causality:
Links to Reprogramming and Neoplasia:

A Brief History of Activator RNA:

"Ultrastructural Probes of Active DNA Sites, and the RNA Activators of DNA". (PowerPoint Presentation).