Science in Society Archive

Epigenetics and Beyond

Epigenetic Inheritance through Sperm Cells, the Lamarckian Dimension in Evolution

New findings on the molecular mechanisms whereby epigenetic changes acquired during development can be transmitted to the next generation via sperm cells are vindicating Lamarck’s theory of evolution that had been completely eclipsed by Darwin’s followers for over a century Dr. Mae-Wan Ho

Lamarck versus Darwin

For nearly a century, the overwhelming majority of biologists held firmly to the ‘neo-Darwinian’ dictum that organisms are strictly determined by their genetic make-up, which is essentially isolated from the environment, so characteristics acquired during one’s lifetime can never be transmitted to the next generation.

Thus, if one of your parents trained very hard to become an Olympic tennis champion, you would not expect be born with particularly strong arms or be a child prodigy in tennis unless you happened to have inherited the right genes from that parent, and it would not have mattered if he/she trained for the championship or not. That was because genes were supposed to remain constant except for rare random mutations; random in the sense that the mutations bear no relationship to the environment, so your parents’ experience can never influence your life.

Neo-Darwinism is derived from a combination of Darwin’s theory of natural selection and Mendel’s theory of genes determining the characteristics of organisms. Neo-Darwinism is opposed to ‘neo-Lamarckism’, derived from Lamarck, who proposed that non-random epigenetic variations arising from the organism’s experience during development are the stuff of evolution, as they can be transmitted to the next generation       

Notable critics of neo-Darwinism such as British evolutionary and developmental biologist Conrad Waddington had realised that the Lamarckian dimension in evolution cannot be ignored, which was why he put ‘epigenesis’, i.e., development at centre stage. He proposed that the intrinsic dynamics of development - the “epigenetic landscape” - was the real source of novel non-random variations for evolution [1] (see Beyond neo-Darwinism: the Epigenetic Approach to Evolution [2], I-SIS scientific publication, for further details.).

Since then, findings in molecular genetics had made neo-Darwinism increasingly untenable in both development and evolution [3, 4] (see Evolution, I-SIS scientific publications, and Living with the Fluid Genome, I-SIS publication). While the dynamics of development remains one of the hardest problems still in search of a solution, Lamarckian mechanisms abound at the molecular level [4, 5] (Life Beyond the Central Dogma series, SiS 24).

New research reveals that you could be getting, not so much the right genes, but ‘retro-genes’ from your Olympic tennis champion father. That is a very significant finding, because mother’s influence through the egg cell and the womb, has long been recognized as an important part of the environmental (non-genetic) input for development, while the father was supposed to contribute nothing before birth except his genes. Now, it appears that the father’s experience, too, could result in ‘retro-genes’ that might be passed on via the sperm.

Retro-genes delivered via sperm cells

‘Retro-genes’ are generated by the process of reverse transcription, in which transcribed RNA, modified, amplified and tested by the individual’s experience, are ‘back copied’ into complementary DNA (cDNA), constituting new genetic messages that may be delivered in sperm cells to the egg at fertilization [6].

This process has been shown to occur during in vitro fertilization, regardless of whether intact sperm cells are incubated with exogenous DNA or RNA molecules. The reverse-transcribed sequences transferred to embryos at fertilization are propagated in mosaic fashion in the tissues of founder animals, and further transmitted to their offspring, where they are maintained as low-copy number structures (episomes) outside the chromosomes.

It is now widely accepted that sperm cells of virtually all animal species can take up DNA molecules and deliver them to the egg at fertilization; this has been exploited to make genetically modified animals with variable efficiency. But the fate of sperm-bound DNA after delivery to the egg is still unclear, in particular, whether foreign nucleic acids become integrated into the host genome, or remain outside the chromosome.

Lab experiments indicate that non-integrated episomal structures are frequently generated when foreign DNA molecules are directly incubated with intact sperm cells that are then used to fertilize eggs. Integration in the host genome is rare under those conditions, and so far, claimed by a single research group experimenting on swine. The same group also later reported the transmission of non-integrated sequences.

In contrast, integration seems to be favoured with protocols that avoid direct interaction between the exogenous nucleic acid molecules and the sperm membrane, such as wrapping the foreign nucleic acids in membrane lipids, or incubating foreign DNA with sperm cells without membranes followed by microinjection into the eggs.

Sperm-mediated gene transfer depends on reverse transcription

Corrado Spadafora at the University of Rome in Italy has identified most of the factors involved in this sperm-mediated gene transfer (SMGT) [6]. The foreign nucleic acids are taken in and reach the nuclear scaffold of sperm cells, where they are rearranged by nucleases and undergo recombination that eventually leads to integration in the sperm genome. Analysis revealed that integration occurs in one, or only very few, preferred sites in the mouse genome, and is therefore very infrequent. External DNA molecules activate one or more nucleases in the sperm, which heavily degrade the foreign DNA, and eventually also cleave a minor chromatin component. These findings suggest that discrete sites of nuclease sensitivity exist within the otherwise tightly packed chromatin of mature sperm cells that are preferential targets for integrating foreign DNA.

The fraction of mouse sperm chromatin that closely resemble active chromatin of somatic cells in being nuclease-sensitive and has very low level of methylation are also unexpectedly enriched in retrotransposon DNA, among which, the most abundant are reverse transcriptase (RT)-encoding LINE1 sequences. This intriguing finding prompted Spadafora’s research team to incubate mouse sperm cells with foreign RNA molecules, and then search for evidence of reverse transcription into cDNA.

In one of these experiments, they incubated sperm cells with RNA transcribed from a construct expressing a b-galactosidase (b-gal) reporter gene. They then used this to fertilize eggs in vitro, and produced a F0 founder generation, followed by a F1 progeny by normal breeding. Direct PCR analysis of DNA samples from both F0 and F1 animal populations confirmed that b-gal containing cDNAs were generated in sperm, delivered to oocytes, and propagated in mosaic fashion through embryonic development in various tissues of the adult animals and transmitted to the next generation.

Remarkably, these sequences are maintained stably as low-copy number episomes (<1 copy per genome), and inherited in non-Mendelian, mosaic fashion. Most significantly, the expression of the b-gal protein was detected in a variety of tissues in both F0 and F1 animals.

Subsequently, the researchers found that an RT-dependent process is triggered not only when sperms are incubated with RNA, but also when they are exposed to DNA. They incubated sperm cells with a retrotransposing cassette containing the DNA construct with enhanced green fluorescence protein (EGFP) as reporter gene interrupted by a g-globin intron placed in the opposite orientation to that of the EGFP. In order to become expressed, the reporter gene must go through reverse transcription. First the DNA construct interacts with the sperm and is taken up into the nucleus where it is transcribed; the primary RNA is then spliced (to remove the interruption g-globin intron) and finally reverse-transcribed to cDNA containing intact EGFP sequence. Interestingly, only a small proportion of the newly synthesized cDNAs is retained within the sperm, while most of it is released into the incubation medium, and available for further interaction with sperm cells. Eventually, a steady state is reached in which the vast majority of the sperm cells are associated with foreign cDNA as extrachromosomal, low-copy number episomes. These are transcribed and the EGFP reporter gene expressed in various tissues of the adult animals.

Rare integration events may occur in the matrix-bound chromosomal DNA which are nuclease sensitive and integration-prone, as opposed to non-matrix bound chromatin.

These findings are impressive, but they only involve artificial in vitro conditions in which sperms are incubated with foreign DNA or RNA. There is so far no evidence that the same processes would take place in vivo. Or is there?

Epigenetic inheritance in vivo

Another team of researchers led by Minoo Rassoulzadegan at the French National Institute for Health and Medical Research (INSERM) in Nice recently reported on a case of non-Mendelian RNA-mediated inheritance of extra-genomic information in mice [7]. The mutant engineered by a LacZ insertion in the Kit gene coding for tyrosine kinase receptor results in mice that die shortly after birth in the homozygous state, while heterozygotes survive with white patches on the tail and feet. The surprising finding is that some of the offspring of such mice, which inherited two wild-type copies of Kit, still exhibited the white patches characteristic of the mutant animals. Similar results are obtained from mating heterozygous mutants to wild-type mice, regardless of the sex of the parent carrying the mutant gene. The mutant phenotype appearing in homozygous wild-type offspring of heterozygous parents is associated with reduced expression of wild-type Kit mRNA, concomitant with an accumulation of Kit RNA transcripts without poly-A tail (pre-mRNA) of abnormal size in tissues, most easily detected by staining for RNA in the sperm. Microinjection of that RNA into fertilized eggs induced a heritable white tail phenotype.

Thus, phenotypes are not exclusively due to chromosomal genes but depend on information apparently stored in a stable class of RNA molecules, which may depend on RNA-dependent RNA polymerase enzyme.

Commenting on these findings, Spadafora thinks it is possible that the propagation and expansion of RNA goes through a DNA intermediate via reverse transcriptase [6]. Such RT-mediated replication and expansion not only takes place in sperm cells and eggs, but perhaps also in embryos and in differentiated somatic cells, given that RT is now known to operate throughout embryogenesis.

Somatic hyper-mutation hypothesis revisited

The proposal that RNA mediates epigenetic inheritance is not new. Immunologist Ted Steele, currently at Australian National University Canberra, first put forward such a theory in his book, Somatic Selection and Adaptive Evolution: On the Inheritance of Acquired Characters, published in 1981 [8], based on the then highly controversial experimental results demonstrating the inheritance of immunological tolerance through the male line. While some of us cheered from the side lines [4], Steele and his fellow researchers were vehemently dismissed and attacked by the establishment throughout the 1980s and 1990s when they continued pushing back the frontiers [8], and are now in danger of being fully vindicated.

Essentially, Steele and colleagues proposed that immune responses to foreign antigens provoke high rates of mutation (hyper-mutation) in white blood cells through cycles of transcription and error-prone reverse-translation until high affinity antigen-binding antibodies are formed. Cells forming such antibodies are selected by the foreign antigen resulting in the enormous clonal expansion of those cells.

Steele thinks that as the Kit-specific effects in the work of Rassoulzadegan and colleagues are transmitted at least to the second generation, the phenomenon must depend on more than just RNA stability. He is convinced that a reverse transcription and genomic integration step must intervene at some stage to fix the DNA in the germline [9]. In the case of Spadafora’s important work, Steele thinks “The [rare] centromeric integration site Spadafora describes may be peculiar to his type of in vitro uptake system [and may be the organism’s way] to quarantine potential genetic effects of exogenous foreign DNA in seminal fluid.”

But what about the ‘natural’ route, involving somatic RNA/DNA delivered to developing sperm or spermatogonia?” Steele asks. That is the crucial question as far as adaptive evolution is concerned.

For more on the problem of development and evolution in the light of the new findins in epigenetics see [11] Development and Evolution Revisited (ISIS scientific preprint).For more on the problem of development and evolution in the light of the new findins in epigenetics see [11] Development and Evolution Revisited (ISIS scientific preprint).

Article first published 23/03/09


  1. Waddington CH. The Strategy fo the Genes, Allen & Unwin, London, 1957.
  2. Ho MW and Saunders PT. Beyond neo-Darwinism – an epigenetic approach to evolution. J theor Biol 1979, 78, 573-91.
  3. Ho MW. Evolution In Encyclopedia of Coomparative Psychology (G. Greenberg ed.), Garland Publication, New York, 1997.
  4. Ho MW. Living with the Fluid Genome, ISIS/TWN, London/Penang, 2003.
  5. Ho MW. Life after the Central Dogma series, Science in Society 24, 4-13+55, 2004.
  6. Spadafora C. Sperm-mediated ‘reverse’ gene transfer: a role of reverse transcriptase in the generation of new genetic information. Human Reproduction 2008, 23(4), 735-40.
  7. Rasoulzadegan M, Grandjean V, Gounon P, Vicent S, Gillot I, Cuzin F. RNA-mediated non-mendelian inheritance of an epigenetic change in the mouse. Nature 2006, 441, 469-74.
  8. Steele EJ. Somatic Selection and Adaptive Evolution: On the Inheritance of Acquired Characters. 2nd Edition, Revised with an author’s Postscript, University of Chicago Press, Chicago, 1981.
  9. Steele EJ, Lindley RA and Blanden RV. Lamarck’s signature: How retrogenes are changing the natural selection paradigm. Allen & Unwin, Frontiers of Science: Series Editor Paul Davies, Sydney, Australia, 1998.
  10. Steele EJ (personal communication in e-mail message to the author, 23 December 2008).
  11. Ho MW. Development and evolution revisited. In Handbook of Developmental Science, Behavior and Genetics (K. Hood, C. Halpern, G. Greenberg & R. Lerner, eds.), Blackwell Publishing, New York, 2009.

Got something to say about this page? Comment

Comment on this article

Comments may be published. All comments are moderated. Name and email details are required.

Email address:
Your comments:
Anti spam question:
How many legs on a cat?