DNA demethylation and hypermethylation are both required for late nodule development in Medicago.

Plant epigenetic regulations are involved in transposable element silencing, developmental processes and responses to the environment1-7. They often involve modifications of DNA methylation, particularly through the DEMETER (DME) demethylase family and RNA-dependent DNA methylation (RdDM)8. Root nodules host rhizobia that can fix atmospheric nitrogen for the plant's benefit in nitrogen-poor soils. The development of indeterminate nodules, as in Medicago truncatula, involves successive waves of gene activation9-12, control of which raises interesting questions. Using laser capture microdissection (LCM) coupled to RNA-sequencing (SYMbiMICS data11), we previously identified 4,309 genes (termed NDD) activated in the nodule differentiation and nitrogen fixation zones, 36% of which belong to co-regulated genomic regions dubbed symbiotic islands13. We found MtDME to be upregulated in the differentiation zone and required for nodule development, and we identified 474 differentially methylated regions hypomethylated in the nodule by analysing ~2% of the genome4. Here, we coupled LCM and whole-genome bisulfite sequencing for a comprehensive view of DNA methylation, integrated with gene expression at the tissue level. Furthermore, using CRISPR-Cas9 mutagenesis of MtDRM2, we showed the importance of RdDM for CHH hypermethylation and nodule development. We thus proposed a model of DNA methylation dynamics during nodule development.

Inhibition of imatinib-mediated apoptosis by the caspase-cleaved form of the tyrosine kinase Lyn in chronic myelogenous leukemia cells.

Once cleaved by caspases, the Lyn tyrosine kinase (LynDeltaN) is relocalized from the plasma membrane to the cytoplasm of apoptotic cells, but the function of such a cleavage is incompletely understood. We evaluated the effect of LynDeltaN overexpression on imatinib sensitivity of the chronic myelogenous leukemia (CML) cell line K562. Therefore, we generated stable cells that express plasmids encoding LynDeltaN or its catalytically inactive counterpart LynDeltaNKD. We established that Lyn is cleaved in imatinib-treated parental K562 cells in a caspase-dependent manner. Lyn cleavage also occurred following BCR-ABL silencing by specific short hairpin RNA (sh-RNA). Imatinib-induced apoptosis was abrogated in LynDeltaN-overexpressing cells, but not in cells overexpressing its inactive counterpart. Conversely, the overexpression of LynDeltaN failed to affect the differentiation of K562 cells. Importantly, the protective effect of LynDeltaN was suppressed by two inhibitors of Lyn activity. LynDeltaN also inhibits imatinib-mediated caspase-3 activation in the small proportion of nilotinib-resistant K562 cells overexpressing Lyn that can engage an apoptotic program upon imatinib stimulation. Finally, Lyn knockdown by sh-RNA altered neither imatinib-mediated apoptosis nor differentiation. Taken together, our data show that the caspase-cleaved form of Lyn exerts a negative feedback on imatinib-mediated CML cell apoptosis that is entirely dependent on its kinase activity and likely on the BCR-ABL pathway.

The presence of GSI-like genes in higher plants: support for the paralogous evolution of GSI and GSII genes.

Glutamine synthetase type I (GSI) genes have previously been described only in prokaryotes except that the fungus Emericella nidulans contains a gene (fluG) which encodes a protein with a large N-terminal domain linked to a C-terminal GSI-like domain. Eukaryotes generally contain the type II (GSII) genes which have been shown to occur also in some prokaryotes. The question of whether GSI and GSII genes are orthologues or paralogues remains a point of controversy. In this article we show that GSI-like genes are widespread in higher plants and have characterized one of the genes from the legume Medicago truncatula. This gene is part of a small gene family and is expressed in many organs of the plant. It encodes a protein similar in size and with between 36 and 46% amino acid sequence similarity to prokaryotic GS proteins used in the analyses, whereas it is larger and with less than 25% similarity to GSII proteins, including those from the same plant species. Phylogenetic analyses suggest that this protein is most similar to putative proteins encoded by expressed sequence tags of other higher plant species (including dicots and a monocot) and forms a cluster with FluG as the most divergent of the GSI sequences. The discovery of GSI-like genes in higher plants supports the paralogous evolution of GSI and GSII genes, which has implications for the use of GS in molecular studies on evolution.

The early nodulin gene MtN6 is a novel marker for events preceding infection of Medicago truncatula roots by Sinorhizobium meliloti.

MtN6 belongs to a series of cDNA clones representing Medicago truncatula genes transcriptionally activated during nodulation by Sinorhizobium meliloti (P. Gamas, F. de Carvalho Niebel, N. Lescure, and J. V. Cullimore, Mol. Plant-Microbe Interact. 9:233-242, 1996). We show here by in situ hybridization that MtN6 transcripts specifically accumulate first at very localized regions in the outer root cell layers, corresponding to outer cortical cells containing preinfection threads. At later stages, MtN6 expression is observed ahead of growing infection threads, including in the infection zone of mature root nodules. Interestingly, regulation of MtN6 is clearly distinct from that of other early nodulins expressed in the same region of the nodule, in terms of response to bacterial symbiotic mutants and to purified Nod factors. We thus suggest that MtN6 represents the first specific marker of a pathway involved in preparation to infection, which is at least partly controlled by Nod factors. Finally, we discuss the intriguing sequence homology shown by MtN6 to a protein from Emericella (Aspergillus) nidulans, FluG, that plays a key role in controlling the organogenesis of conidiophores (B. N. Lee and T. H. Adams, Genes Dev. 8:641-651, 1994).

The Medicago truncatula MtAnn1 gene encoding an annexin is induced by Nod factors and during the symbiotic interaction with Rhizobium meliloti.

Here we report the characterization of a new Nod factor-induced gene from Medicago truncatula identified by mRNA differential display. This gene, designated MtAnn1, encodes a protein homologous to the annexin family of calcium- and phospholipid-binding proteins. We further show that the MtAnn1 gene is also induced during symbiotic associations with Rhizobium meliloti, both at early stages in bacterial-inoculated roots and in nodule structures. By in situ hybridization, we demonstrate that MtAnn1 expression in nodules is mainly associated with the distal region of invasion zone II not containing infection threads, revealing MtAnn1 as a new marker gene of the pre-infection zone. Moreover, analyses of MtAnn1 expression in response to bacterial symbiotic mutants suggest that the expression of MtAnn1 during nodulation requires biologically active Nod factors and is independent of the infection process.

Symbiosis-specific expression of two Medicago truncatula nodulin genes, MtN1 and MtN13, encoding products homologous to plant defense proteins.

Two Medicago truncatula nodulin genes putatively encoding proteins structurally related to two classes of proteins commonly associated with plant defense reactions have been characterized. MtN1 is homologous to two small, cysteine-rich, pathogen-inducible proteins from pea (pI39 and pI230), whereas MtN13 is closely related to the PR10 family of pathogenesis-related proteins. We show that neither MtN1 nor MtN13 is induced in leaves in response to pathogenic bacteria, and that both are exclusively expressed during nodulation. In situ hybridization experiments as well as Northern (RNA) studies of interactions between M. truncatula and either wild-type Rhizobium meliloti or mutants deficient in infection establish that MtN1 is associated with the infection process, while MtN13 represents the first specific marker described for the nodule outer cortex. Possible roles for MtN1 and MtN13 are discussed. We also present the identification of another member of the PR10 family, designated as MtPR10-1, whose regulation is strikingly different from that observed for MtN13, being constitutively expressed in roots and pathogen-inducible in leaves.

Use of a subtractive hybridization approach to identify new Medicago truncatula genes induced during root nodule development.

We report the identification of new molecular markers associated with different stages of Rhizobium-induced nodule development in the legume Medicago truncatula. A cDNA library was constructed from pre-nitrogen-fixing M. Truncatula nodules, and differentially screened with a polymerase chain reaction-amplified subtracted probe. Twenty-nine new families of nodulin cDNA clones, designated MtN1 to MtN29, were thus identified in addition to clones for several known nodulins. All MtN genes were shown by Northern (RNA) hybridization analysis to be induced during nodulation, some of them well before nodule emergence. The MtN genes were classified into three groups depending on their expression kinetics. The expression of three MtN genes showed a limited induction by Nod factors purified from Rhizobium meliloti. Homologies with a variety of proteins were found for the deduced amino acid sequences of 10 of the MtN genes.

Purification and characterization of TnsC, a Tn7 transposition protein that binds ATP and DNA.

The bacterial transposon Tn7 encodes five transposition genes tnsABCDE. We report a simple and rapid procedure for the purification of TnsC protein. We show that purified TnsC is active in and required for Tn7 transposition in a cell-free recombination system. This finding demonstrates that TnsC participates directly in Tn7 transposition and explains the requirement for tnsC function in Tn7 transposition. We have found that TnsC binds adenine nucleotides and is thus a likely site of action of the essential ATP cofactor in Tn7 transposition. We also report that TnsC binds non-specifically to DNA in the presence of ATP or the generally non-hydrolyzable analogues AMP-PNP and ATP-gamma-S, and that TnsC displays little affinity for DNA in the presence of ADP. We speculate that TnsC plays a central role in the selection of target DNA during Tn7 transposition.

Tn7 transposition in vitro proceeds through an excised transposon intermediate generated by staggered breaks in DNA.

We have developed a cell-free system in which the bacterial transposon Tn7 inserts at high frequency into its preferred target site in the Escherichia coli chromosome, attTn7; Tn7 transposition in vitro requires ATP and Tn7-encoded proteins. Tn7 transposes via a cut and paste mechanism in which the element is excised from the donor DNA by staggered double-strand breaks and then inserted into attTn7 by the joining of 3' transposon ends to 5' target ends. Neither recombination intermediates nor products are observed in the absence of any protein component or DNA substrate. Thus, we suggest that Tn7 transposition occurs in a nucleoprotein complex containing several proteins and the substrate DNAs and that recognition of attTn7 within this complex provokes strand cleavages at the Tn7 ends.

Functional organization of the ends of IS1: specific binding site for an IS 1-encoded protein.

The IS 1-encoded protein InsA binds specifically to both ends of IS1, and acts as a repressor of IS1 gene expression and may be a direct inhibitor of the transposition process. We show here, using DNasel 'foot-printing' and gel retardation, that the InsA binding sites are located within the 24/25 bp minimal active ends of IS1 and that InsA induces DNA bending upon binding. Conformational modification of the ends of IS1 as a result of binding of the host protein integration host factor (IHF) to its site within the minimal ends has been previously observed. Using a collection of synthetic mutant ends we have mapped some of the nucleotide sequence requirements for InsA binding and for transposition activity. We show that sequences necessary for InsA binding are also essential for transposition activity. We demonstrate that InsA and IHF binding sites overlap since some sequence determinants are shared by both InsA and IHF. The data suggest that these ends contain two functional domains: one for binding of InsA and IHF, and the other for transposition activity. A third region, when present, may enhance transposition activity with an intact right end. This 'architecture' of the ends of IS1 is remarkably similar to that of IS elements IS10, IS50 and IS903.

Functional organization of the ends of IS 1: specific binding site for an IS1-encoded protein.

The IS 1-encoded protein InsA binds specifically to both ends of IS 1, and acts as a repressor of IS1 gene expression and may be a direct inhibitor of the transposition process. We show here, using DNasel 'foot-printing' and gel retardation, that the InsA binding sites are located within the 24/25bp minimal active ends of IS1 and that InsA induces DNA bending upon binding. Conformational modification of the ends of IS 1 as a result of binding of the host protein integration host factor (IHF) to its site within the minimal ends has been previously observed. Using a collection of synthetic mutant ends we have mapped some of the nucleotide sequence requirements for InsA binding and for transposition activity. We show that sequences necessary for InsA binding are also essential for transposition activity. We demonstrate that InsA and IHF binding sites overlap since some sequence determinants are shared by both InsA and IHF. The data suggest that these ends contain two functional domains: one for binding of InsA and IHF, and the other for transposition activity. A third region, when present, may enhance transposition activity with an intact right end. This 'architecture' of the ends of IS 1 is remarkably similar to that of IS elements IS10, IS50 and IS903.

Escherichia coli integration host factor binds specifically to the ends of the insertion sequence IS1 and to its major insertion hot-spot in pBR322.

We report here that the ends of IS1 are bound and protected in vitro by the heterodimeric protein integration host factor (IHF). Under identical conditions, RNA polymerase binds to one of these ends (IRL) and protects a region that includes the sequences protected by IHF. Other potential sites within IS1, identified by their homology to the apparent consensus sequence, are not protected. Footprinting analysis of deletion derivatives of the ends demonstrates a correspondence between the ability of the end sequence to bind IHF and its ability to function as an end in transposition. Nonetheless, some transposition occurs in IHF- cells, indicating that IHF is not an essential component of the transposition apparatus. IHF also binds and protects four closely spaced regions within the major hot-spot for insertion of IS1 in the plasmid pBR322. This striking correlation of hot-spot and IHF-binding sites suggests a possible role for IHF in IS1 insertion specificity.

Expression of F transfer functions depends on the Escherichia coli integration host factor.

We present evidence that the Escherichia coli DNA binding protein, IHF, plays an important role in conjugal transfer of the plasmid F. Our results suggest that IHF exerts this effect by positively effecting transcription of the transfer (tra) operon of the plasmid.

Artificial transposable elements in the study of the ends of IS1.

We have constructed artificial IS1-based transposons by attaching synthetic oligodeoxynucleotides, corresponding to the sequence of the ends of IS1, to a selectable DNA segment ['omega' fragment; Prentki and Krisch, Gene 29 (1984) 303-313]. These transposons were used to examine the sequence requirements at the ends for IS1 transposition. We show here that a 24- to 28-bp sequence from the left or right ends of IS1 is capable of transposition when present at both ends of the omega fragment in the correct orientation. Transposition activity requires the presence of an intact IS1 in cis on the same plasmid molecule. In trans, however, neither resident genomic copies of IS1, nor copies carried by a compatible, high-copy-number plasmid present in the same cell, complement the artificial transposons efficiently. Transposition frequencies in the presence of a cis-complementing IS1 are, however, similar to those of the naturally occurring IS1-based transposon, Tn9. In addition, transposition results in a 9-bp duplication in the target DNA molecule as is usually the case for insertion of the intact IS1. Using this system, we have obtained evidence indicating that the activity of a synthetic IS1 end is not determined exclusively by its sequence, but can be strongly enhanced by a second, wild-type end used in the transposition event. The data also show that single base pair mutations can exhibit a cumulative effect in reducing transposition activity.

Replication of pSC101: effects of mutations in the E. coli DNA binding protein IHF.

We have shown that the plasmid pSC101 is unable to be maintained in strains of E. coli carrying deletions in the genes himA and hip which specify the pleitropic heterodimeric DNA binding protein, IHF. We show that this effect is not due to a modulation of the expression of the pSC101 RepA protein, required for replication of the plasmid. Inspection of the DNA sequence of the essential replication region of pSC101 reveals the presence of a site, located between the DnaA binding-site and that of RepA, which shows extensive homology with the consensus IHF binding site. The proximity of the sites suggests that these three proteins, IHF, DnaA, and RepA may interact in generating a specific DNA structure required for initiation of pSC101 replication.

DNA sequence at the end of IS1 required for transposition.

The insertion sequence IS1 belongs to a class of bacterial transposable genetic elements that can form compound transposons in which two copies of IS1 flank an otherwise non-transposable segment of DNA. IS1 differs from other known elements of this class (such as IS10, IS50 and IS903) in several respects. It is one of the smallest known insertion elements, exhibits a relatively complex array of open reading frames, is present in the chromosomes of various Enterobacteria, in some cases in many copies, and its insertion can result in the duplication of either 8 or 9 base pairs (bp) in the target DNA. Furthermore, although, like other members of the compound class, it seems to undergo direct transposition, IS1 also promotes replicon fusion (co-integrate formation) at a relatively high frequency. Like all other elements studied to date, the integrity of the extremities of IS1 are essential for efficient transposition. We have constructed a test system to determine the minimal DNA sequences at the extremities of IS1 required for transposition. Sequential deletions of the end sequences reveal that 21-25 bp of an isolated extremity are sufficient for transposition. A specific sequence 13-23 bp from the ends, defining the edge of the minimal sequence, is implicated as an essential site. The sites, symmetrically arrayed at both ends of IS1, correspond to the apparent consensus sequence of the known binding sites for the Escherichia coli DNA-binding protein (called integration host factor or IHF) which is required for the site-specific recombination that leads to integration of bacteriophage lambda into the bacterial genome. The sites at the ends of IS1 may thus bind a host protein, such as JHF or a related protein, that is involved in regulating the transposition apparatus.

Specificity of insertion of IS1.

A systematic study of the specificity of insertion of the transposable element IS1 into small defined-sequence plasmids (pBR322 and derivatives) was conducted to determine the features of the DNA sequence that influence target site selection. We have physically mapped several collections of independent insertions of IS1 into these plasmids and have determined: (1) that about 80% of all insertions occur in the DNA segment (about 200 base-pairs) between the unique EcoRI site of pBR322 and the beginning of the beta-lactamase gene, one of the two regions of high A + T density in this plasmid; (2) that there is a strong orientation effect in this region (almost all IS1 insertions are in one orientation) that depends on both the pBR322 sequence and the environment of the transposon in the donor molecule; and (3) that the orientation effect does not depend on the strong transcription that is directed through this region in pBR322. Furthermore, we have found that insertion of a poly(dA X dT) segment into pBR322 creates an artificial hotspot for IS1 insertion, even though it is not as attractive for insertion as the above-mentioned major hotspot. Our observations suggest that an interplay between several properties of the target sequences and the sequence environment of the donor transposon is responsible for the observed specificity of position and orientation. One of the possibilities discussed here is that preferred "entry-sites", or "signal" sequences, for the transposition complex play a major role in determining the positions and orientations of IS1 insertions.