A hitchhiker’s guide to foreign genomes

In 1928, bacteriologist Frederick Griffith performed experiments with pneumococcal bacteria in mice. He found that avirulent pneumococci became virulent when they were in contact with a virulent strain (Griffith, 1928). It was the first experiment that showed that genetic material could be transferred between organisms other than from parent to offspring. Nowadays, this mechanism – called horizontal gene transfer – is known as a major driver of evolution in bacteria. For a long time, horizontal gene transfer was thought to be limited to prokaryotes. Reports on the mechanism in eukaryotes were met with skepticism. The criticism was not completely unfounded: proving horizontal gene transfer in eukaryotes is difficult. Eukaryotic organisms are often associated with microbes, which causes contamination of samples with bacterial DNA fragments. In the last few years, it became clear that such contaminations led to false detection of horizontal gene transfer (e.g., Richards and Monier, 2016). However, it is gradually becoming evident that horizontal gene transfer does exist in eukaryotic organisms, including plants. Nevertheless, the reports in plants were limited (reviewed by Wickell and Li, 2020). Most reports on horizontal gene transfer in plants come from parasitic plants. Close physical association between two plant species, e.g., between a parasitic plant and its host, is thought to promote exchange of genetic material. Another mechanism by which plants might exchange genes is with a vector as a bridge, such as fungi or viruses. However, the exact pathways of horizontal gene transfer in plants are not well understood (Gao et al., 2014). Several years ago, plant researchers stumbled upon something peculiar. The genomes of wild barley (Hordeum) species contained alien DNA sequences that matched sequences from Panicum (Mahelka and Kopecký, 2010). It was a lucky find. The researchers were investigating a different grass, Elymus repens, which is an allohexaploid species. To understand which parental species contributed to its genome, the researchers screened for ribosomal DNA (rDNA) genes, which are used as sequence-based markers to untangle polyploidy. They found the Panicum-like rDNA sequence in the subgenome of E. repens that is derived from wild barley. The researchers were intrigued because the lineages to which wild barley and Panicum belong (pooid and panicoid grasses) had split about 50 million years ago (Kumar et al., 2017). Recently, they established that the foreign DNA is indeed present in wild barley species, and was transferred by horizontal gene transfer (Mahelka et al., 2017). In this issue, they managed to characterize the complete DNA fragment and found that it contains more than rDNA genes (Mahelka et al., 2021). To characterize the complete fragment, the authors used bacterial artificial chromosome (BAC) libraries constructed from flow-sorted nuclei. In this way, they could ensure that any observed foreign DNA was truly incorporated in the nuclear genome, and not caused by DNA contamination. They constructed libraries of two different wild barley species, i.e., the Asian species Hordeum bogdanii and the South American species Hordeum pubiflorum. By screening the libraries with probes that targeted Panicum-like DNA, they found several BAC clones that harbored panicoid DNA. Fluorescence in situ hybridization and sequencing of the BAC clones showed that the different positive hits represent a single chromosomal fragment in both species. This fragment consists of repeated blocks of DNA that showed high similarity between the two species, which indicates a common origin. The researchers analyzed the sequence of the foreign DNA and discovered that besides rDNA genes, the fragment also contains transposable elements (Figure). Most of these transposons belonged to the group of long terminal repeat (LTR) retrotransposons. This was a useful finding because the LTRs in these transposons can be used for molecular dating. By analyzing the divergence of the LTR elements, the researchers were able to estimate the insertion time of the retrotransposons between 2.9 and 1.7 million years ago. Keeping in mind that the retrotransposons could have remained active for some time after the transfer, this matches the estimation based on phylogeny that the fragment was inserted between 5 and 1.7 million years ago. This is long after the split of the pooid and panicoid lineages. Hordeum (wild barley) species carry a chromosomal fragment that was obtained from Panicum. Whereas the lineages of pooid and panicoid grasses split about 50 million years ago, the Panicum-like DNA fragment in Hordeum was obtained by horizontal gene transfer between 2.9 and 1.7 million years ago. The fragment contains protein-coding genes, ribosomal DNA genes (rDNA) and transposable elements (TEs). Before integration in Hordeum, different Panicum-like DNA fragments were rearranged into a single locus. At least one of the genes is potentially functional. Figure by Václav Mahelka (created with BioRender.com). In addition to the rDNA genes and transposable elements, the authors identified five protein-coding genes on the foreign DNA fragment: ABC transporter B family member 1; RNA polymerase sigma factor sigB; DEAD-box ATP-dependent RNA helicase 39; Ervatamin-C-like; and glutathione S-transferase T3-like (Figure). They wanted to know if these genes were functional, and tested if they were under purifying selection. Purifying selection removes deleterious mutations and is therefore a sign that a gene is functional. A good way to infer selection pressure is the ratio between the non-synonymous and synonymous substitution rate. In all five genes, they found more synonymous than non-synonymous substitutions, which indicates purifying selection. However, closer inspection of the gene sequences showed that most of them did not comprise a complete set of exons, or contained premature stop codons. Based on these observations, only the Ervatamin-C-like sequences might be functional. This was supported by the finding that when they analyzed expression of the five genes, only Ervatamin-C-like was consistently expressed. Interestingly, the Ervatamin-C-like gene does not have any native homologs in wild barley. Thus, when the foreign fragment was introduced, Ervatamin-C-like was a new gene for the gene pool. Ervatamins are cysteine proteases that are potentially involved in protein maturation, degradation and rebuilding in response to different external stimuli. It remains enigmatic how the panicoid fragment was transferred into wild barley. The researchers noticed a short fragment similar to rice tungro bacilliform virus in the Panicum-like fragment in H. bogdanii. However, they were unable to identify if the fragment got into the genome of H. bogdanii from another plant species, or directly from the virus. In fact, it is not clear if horizontal gene transfer in grasses needs to be mediated by a vector. Therefore, the corresponding authors Václav Mahelka and Jan Šafář are trying to understand how genetic material is transferred in grasses. The outcomes might help us understand if horizontal gene transfer is more common in plants than we currently think.