Le cytomégalovirus humain (CMVH) est un pathogène opportuniste majeur en cas d’immunodépression et représente la principale cause d’infection congénitale d’origine virale. Bien qu’efficaces, l’utilisation des molécules conventionnelles est limitée par l’émergence de résistance et leur toxicité. Il devient alors nécessaire de développer de nouveaux traitements.L’étude des nouvelles mutations émergeant sous traitement antiviral demeure donc essentielle. L’introduction de ces nouvelles mutations, par mutagénèse « en passant », dans un chromosome bactérien artificiel contenant le génome viral nous permet, après transfection en cellules humaines, de tester la sensibilité de la souche recombinantes aux antiviraux.Différentes mutations de résistances ont ainsi été caractérisées. Afin de mettre en évidence de nouvelles cibles antivirales, des analyses bio-informatiques et la production de virus recombinants ont permis d’identifier de potentiels motifs fonctionnels essentiels à la réplication au sein du complexe terminase et hélicase-primase. Ainsi, nous avons montré quela sous-unité pUL56 du complexe terminase appartient à la famille des LAGLIDADG Homing Endonuclease. En effet, pUL56 contient un motif LATLNDIERFL et un motif de liaison à l’ADN. La technologie Alpha utilisant des protéines purifiées a permis de valider le caractère essentiel du fragment WMVVKYMGFF de pUL56 pour l’interaction avec pUL89. Enfin, nous avons mis en évidence les résidus impliqués dans la fixation de l’ATP au sein de l’hélicase et dans la stabilisation du zinc de la primase. Ainsi, la compréhension de la structure de ces protéines pourrait permettent de mieux appréhender leur fonctionnement au sein du processus de réplication du CMVH et le développement de nouvelles thérapies ciblant ces domaines. / Human cytomegalovirus (HCMV) is an important opportunistic pathogen for immunecompromised patients and is the leading cause of congenital viral infection. Although they are effective, using of conventional molecules is limited by the emergence of resistance and their toxicity. Then it becomes necessary to develop new treatments. Study of new mutationsemerging under antiviral treatment is therefore essential. Introduction of these new mutations, by « en passant » mutagenesis, into an artificial bacterial chromosome containing the viral genome allows us, after transfection into human cells, testing antivirals sensitivity of the recombinant. Different mutations of resistances have been characterized. In order tohighlight new antiviral targets, bioinformatics and recombinant viruses production allowed to identify potential functional patterns essential for viral replication within terminase and helicase-primase complex. Thus, we have shown that pUL56 subunit of the terminase complex belongs to the LAGLIDADG Homing Endonuclease family. Indeed, pUL56 contains aLATLNDIERFL motif and a DNA binding motif. Alpha technology using purified proteins allowed to validate the essential character of the WMVVKYMGFF fragment of pUL56 for the interaction with pUL89. Finally, we highlighted the residues involved in ATP binding within the helicase and in the stabilization of zinc within the primase. Thus, understanding of these proteins structure could allow us to better understand their role within the viral replication process and the development of new therapies targeting these domains.

La reconnaissance spécifique du génome viral et son encapsidation est une étape cruciale pour l’assemblage de particules virales. Chez SPP1, comme chez d’autres bactériophages à queue, le moteur moléculaire qui encapside le génome viral est composé de la terminase, une enzyme hétéro-oligomérique qui possède une activité ATPasique et nucléasique, et de la protéine portale, un oligomère cyclique par lequel l’ADN viral est transloqué. Dans un grand nombre de ses virus, l’encapsidation de l’ADN est initiée par la reconnaissance et le clivage d’une séquence spécifique nommée « pac ». C’est un évènement qui se produit une seule fois au début d’une série de cycles d’encapsidation processive à partir d’un concatémère issu de la réplication du génome du phage. La région pac de SPP1 contient deux séquences (pacL et pacR) où TerS (gp1) se lie entourant la région (pacC) où TerL (gp2) coupe l’ADN de SPP1.Ici, nous montrons qu’une région de la séquence pacL et qu’un motif polyadénine de pacR agissent ensemble pour promouvoir le clivage en pacC. La dégénération de la région pacC n’a pas montré d’effet sur que le clivage endonucléolytique qui a lieu à une position bien définie de pacC avec une précision de ~6 pb. Des études avec des phages proches de SPP1 ont montré une conservation dans la position du clivage, malgré des variations dans pacC, pacR ou dans la distance entre pacL et pacC. Les données sont compatibles avec un modèle dans lequel TerS interagit spécifiquement avec la région pacL, sur laquelle le multimère cyclique TerS doit s’enrouler, et le motif polyadénine de la région pacR. Le complexe nucléoprotéique résultant va créer un contexte structural qui permet de recruter et positionner le domaine nucléase de TerL pour une coupure très précise sur pacC sans spécificité de séquence. / The specific recognition of the viral genome and its packaging is a critical step in viral particle assembly. In SPP1, as in many tailed bacteriophages, the macromolecular motor that encapsidates viral DNA is composed of terminase, a hetero-oligomeric enzyme possessing ATPase and nuclease activities, and of portal protein, a cyclic oligomer through which DNA is translocated. In a large number of these viruses, DNA packaging is initiated by recognition and cleavage of a specific sequence pac. This event occurs once at the beginning of a series of processive encapsidation events along a substrate concatemer of replicated phage genomes. The SPP1 pac region has two sequences where TerS (gp1) binds (pacL and pacR) flanking the segment where TerL (gp2) cleaves the SPP1 DNA (pacC). Here we show that a sequence segment of pacL and a poly-adenine motif in pacR act together to promote cleavage at pacC. Extensive degeneration of pacC sequence has no detectable effect in pac cleavage. The endonucleolytic cut occurs at a defined position with a precision of ~6 bp. Studies with SPP1-related phages show conservation of the cut position, irrespectively of sequence variation in pacC, in pacR or changes in pacL-pacC distance. The data is compatible with a model in which TerS interacts specifically with a region of pacL that probably wraps around the TerS cyclical multimer, and a poly-A tract in pacR. The resulting nucleoprotein complex architecture positions TerL for accurate cleavage at pacC without specific sequence requirement.

Die Verpackung und Spaltung konkatemerer DNA ist ein essentieller Prozeß bei der Reifung von Virionen. Die an diesem Prozess maßgeblich beteiligten Proteine werden als Terminase bezeichnet. Die Inhibition der HCMV Terminase bietet einen attraktiven alternativen Ansatzpunkt für die antivirale Therapie. Zur Inhibition der Terminase Aktivität wurden die neuen Benzimidazol D-Ribonukleosid Derivate BTCRB und Cl4RB auf ihre antivirale Wirkung analysiert. Die neuen Benzimidazol D-Ribonukleosid Derivate BTCRB und Cl4RB zeigten sowohl gegen den HCMV Laborstamm AD169 als auch gegen klinische HCMV-Isolaten eine Wirkung. Weiterhin wurde die Wirksamkeit der Substanzen auf anderen Herpesviren getestet. Interessanterweise zeigten BTCRB und Cl4RB sowohl einen Effekt gegen Varicella-Zoster-Virus (VZV) als auch Ratten-Cytomegalovirus (RCMV), wohingegen der Effekt gegen Herpes-simplex-Virus Typ-1 (HSV-1) und Maus-Cytomegalovirus (MCMV) gering war. Infolgedessen eignen sich beide Substanzen BTCRB und Cl4RB als attraktive alternative Inhibitoren für die weitere Entwicklung einer antiviralen Therapie. / DNA packaging is the key step in viral maturation and involves binding and cleavage of viral DNA containing specific DNA-packaging motifs. This process is mediated by a group of specific enzymes called terminase. The development for an inhibitor of HCMV terminase would be of great value, because it would act subsequent to DNA synthesis and block the first steps in viral maturation. Therefore we characterized two inhibitors targeting the HCMV terminase, 2-bromo-4,5,6-trichloro-1-(2,3,5-tri-O-acetly-ß-D-ribofuranosyl) benzimidazole (BTCRB) and 2,4,5,6-tetrachloro-1-(2,3,5-tri-O-acetly-ß-D-ribofuranosyl) benzimidazole (Cl4RB). By using viral plaque formation, viral yield, viral growth kinetics and electron microscopy, we demonstrated that two compounds BTCRB und Cl4RB are highly active against HCMV AD 169 and HCMV clinical isolates. In addition, the antiviral effect on other herpesviruses was determined. Interestingly BTCRB was active all tested herpesviruses. The best effects were observed on VZV-and RCMV-infected cells. Therefore new compounds might be promising attractive compounds for antiviral therapy in the future.

Der DNA-Verpackungsmechanismus des humanen Cytomegalievirus (HCMV) ist charakteristisch für große DNA-Viren wie Herpesviren und ds-Bakteriophagen. Er beruht auf der Spaltung der konkatemeren DNA durch einen viralen, hetero-oligomeren Proteinkomplex, der Terminase. In der vorliegenden Arbeit konnten die funktionellen Domänen der Terminase-Untereinheit pUL89 in vitro identifiziert und charakterisiert werden. Neben einer Nuklease-Aktivität besitzt pUL89 auch die Fähigkeit dsDNA sequenz-unabhängig zu binden. Durch Nuklease-Untersuchungen konnte gezeigt werden, dass pUL89 sowohl dsDNA, als auch lineare DNA spaltet. PUL89 weist dabei eine größere Spezifität zu dsDNA auf. Des Weiteren konnte nachgewiesen werden, dass die Aminosäure D463 eine zentrale Funktion innerhalb der Nuklease-Aktivität besitzt. Durch kolorimetrische DNA-Bindungsuntersuchungen konnte die Aminosäure R544 als essenziell für die dsDNA-Bindungsfähigkeit von pUL89 identifiziert werden. Basierend auf den in vitro Ergebnissen wurden rekombinanten TB40/E-Virusmutanten mit Mutationen im ORF UL89 durch die En Passant Mutagenese generiert. Mit Hilfe dieser Viren sollte der Einfluss der Mutationen auf die Replikation des Virus charakterisiert werden. Es war möglich nachzuweisen, dass die Aminosäuren E534 und R544 eine essenzielle Aufgabe innerhalb von HCMV erfüllen, da die Mutation einer dieser Aminosäure zu nicht wachstums-fähigen BAC-Mutanten führte. Zur Charakterisierung dieser Konstrukte wurden die Zelllinien HELF Fi301-UL89 und HELF Fi301-vProm-UL89 verwendet. Durch Untersuchungen hinsichtlich der Wachstumseigenschaften, Proteinexpression, DNA-Spaltung, DNA-Bindung sowie elektronenmikroskopischen Aufnahmen, konnte gezeigt werden, dass die wachstums-kompetenten BAC-Mutanten keinen signifikanten Unterschied zum Wildtyp-Virus TB40/E zeigten. Sodass nachgewiesen werden konnte, dass die basischen Aminosäuren H565 und H571 keine essenzielle Funktion in pUL89 erfüllen. / The human cytomegalovirus DNA packaging mechanism is characteristic for large DNA viruses like Herpes viruses and ds bacteriophages. This mechanism is based on the cleavage of concatemeric DNA by the viral heterooligomeric protein complex terminase. This dissertation includes the identification and characterization of functional domains of the HCMV terminase subunit pUL89. PUL89 contain a nuclease activity and the ability to bind dsDNA. This protein shows the property to cut as well dsDNA as linear DNA. The amino acid D463 shows a significant role in this cleavage event. Colorimetric DNA binding experiments show the central role of R544 in DNA binding by pUL89. Based of the in vitro results recombinant TB40/E viruses with mutations in the ORF UL89 were generated. These viruses allow a characterization of the impact of virus replication. It was possible to show that the amino acids E534 and R544 have a functional role in HCMV. The mutation of one of these amino acids was enough to generate a growth deficient mutant. The stable cell lines HELF Fi301-UL89 and HELF Fi301-vProm UL89 were used for the characterization of the growth deficient mutants. The growth competent mutants H656A and H571A show no significant differences in comparison with the wild type TB40/E virus. This was verified by growth kinetics, protein expression characterizations, pulse field gel electrophoresis, DNA binding assays and electron microscopy.

Current FDA approved anti-cytomegalovirus antivirals are ganciclovir, foscarnet, and cidofovir. These drugs target the viral polymerase to inhibit DNA synthesis. Halogenated benzimidazoles target a step later in viral replication during packaging and cleavage of the viral genome. The compounds 2-Bromo-5,6-dichloro-1-(β-D-ribofuranosyl)benzimidazole (BDCRB) and 2,5,6-trichloro-1-(β-D-ribofuranosyl)benzimidazole (TCRB) are novel inhibitors of human cytomegalovirus (HCMV) and resistance to these compounds are found within the human cytomegalovirus viral terminase. Terminase is unique to the virus and in theory provides a good target for antiviral development. Beginning with a BDCRB resistant guinea pig cytomegalovirus observations found a single mutation located in the viral terminase gene UL89 and unique formation found in genomic ends. In guinea pigs, the virus continued to produce large amounts of monomer. This is in contrast to human cytomegalovirus treated with either BDCRB or TCRB. Both compounds produced very little monomer DNA but created a supergenomic species, called monomer-plus that migrates to the apparent size of 270-kb on a pulse field gel. A model was developed to explain formation of monomer plus and my project aims originated from the model. In the presence of BDCRB and TCRB, packaging is relaxed resulting in the normal cleavage site being skipped and cleavage occurring at the next available cleavage site creating a short-long-short genome. In guinea pigs, cleavage has been relaxed by premature cleavage of the terminal ends. Current antivirals do not block viral entry therefore, patients are infected with HCMV. Blocking entry of HCMV into cells would prevent HCMV infection. It has been shown that antibodies to epitopes within the gH/gL/UL128-131 complex can block viral entry into endothelial, epithelial, and other cell types while antibodies to gB block entry into fibroblasts. The majority of the current work on entry into epithelial cells has been performed in retinal pigment epithelium and epithelial cells derived from tumors. Viral entry into cell lines derived from mucosa cells was dependent on the complex gH/gL/UL128-131. Rabbit antibodies raised against UL130 and UL131 peptides neutralized epithelial entry with effects as potent as human seropositive sera. This suggests that the entry complex can be blocked by single epitopes.

Staphylococcus aureus pathogenicity island SaPI1 is a genomic element that is mobilized and transduced at high frequency by helper phage 80α. SaPI1 encodes a small terminase protein that belongs to the phage small terminase subunit family. The presence of SaPI1-encoded small terminase suggests that it plays a role in SaPI1-specific packaging into transducing particles by complexing with the 80α large terminase subunit and redirecting recognition to a pac site on SaPI1 DNA from 80α DNA. The effects of deleting the small terminase genes in SaPI1 and in a prophage copy of 80α are consistent with this hypothesis. Induction of the 80α small terminase deletion mutant produces wild type levels of SaPI1 transducing particles, demonstrating that SaPI1 small terminase can replace that of 80a in SaPI1 packaging. Southern blot analysis of virion DNAs isolated from the deletion mutants confirms that SaPI1 redirects packaging of its DNA into SaPI1-sized capsids.

Many double-stranded DNA viruses use a packaging motor during maturation to recognize and transport genetic material into the capsid. In terminase motors, the TerS complex recognizes DNA, while the TerL motor packages the DNA into the capsid shell. Although there are several models for DNA recognition and translocation, how the motor components assemble and power DNA translocation is unknown. Using the thermophilic P74-26 bacteriophage model system, we discover that TerL uses a trans-activated ATP hydrolysis mechanism. Additionally, we identify critical residues for TerL ATP hydrolysis and DNA binding. With a combination of x-ray crystallography, SAXS, and molecular docking, we build a structural model for TerL pentamer assembly. Apo and ATP analog-bound TerL ATPase domain crystal structures show ligand-dependent conformational changes, which we propose power DNA translocation. Together, we assimilate these findings to build models for both motor assembly and DNA translocation. Additionally, with the P76-26 system, we identify the TerS protein as gp83. I find that P74-26 TerS is a nonameric ring that stimulates TerL ATPase activity while inhibiting TerL nuclease activity. Using cryoEM, I solve 3.8 Å and 4.8 Å resolution symmetric and asymmetric reconstructions of the TerS ring. I observe in P74-26 TerS, the conserved C-terminal beta-barrel is absent, and instead the region is flexible or unstructured. Furthermore, the helix-turn-helix motifs of P74-26 TerS are positioned differently than those of known TerS structures, suggesting P74-26 uses an alternative mechanism to recognize DNA.