Open Access

Many viral structural proteins can self-assemble into virus-like particles (VLPs). VLPs can serve as an effective vaccine or be used as a vaccine platform. One of these structural proteins is the hepatitis B virus core antigen (HBcAg), which appears to be suitable as an antigen carrier due to its high immunogenicity. HBcAg has a major immunodominant region (MIR) that is presented on the surface of the VLPs after self-assembly. Foreign antigens can be inserted into this region. Since HBcAg VLPs, unlike the Hepatitis B virus (HBV), do not have an envelope, they are not able to penetrate cell membranes efficiently. As an extracellular antigen, HBcAg VLPs primarily induce a strong humoral immune response. In the present study, we investigated the extent to which HBcAg can be modified to also elicit an enhanced cellular, particularly a cytotoxic, immune response. A cytotoxic CD8+ T cell response is predominantly induced by intracellular antigens. Therefore, our goal was to increase the cell penetration capacity of VLPs. We aimed to achieve this by fusing cell-penetrating peptides (CPPs) to HBcAg. CPPs can spontaneously penetrate cell membranes to enter the cytoplasm of cells. To guarantee that the CCPs were localized to the surface of the VLPs, we fused CPPs to the N-terminus of HBcAg. The CCPs were followed by a tag to allow the purification of VLPs. The T cell epitopes, against which the induced CTL should be directed, were derived from the Large T antigen and inserted into the MIR of HBcAg. Finally, we fused fluorescent proteins to the C-terminus of HBcAg to track the entry of VLPs into cells. Modifications of HBcAg may lead to reduced stability or altered structure of VLPs. To analyze the stability of VLPs, we used nanoscale differential scanning fluorimetry (nanoDSF) analysis. This revealed that the N-terminal fusion of CPPs or the tag to HBcAg does not reduce VLP stability. However, some peptides incorporated into the MIR had a significant effect on the structure and stability of the VLPs. While the incorporation of a Flag-tag or a peptide from ovalbumin had no negative effect on VLP stability, the incorporation of peptides representing T cell epitopes of Large T antigen interfered with VLP formation. Denaturation and reassembly of the aggregates significantly improved the homogeneity of the VLPs, and the C-terminal addition of arginine-rich domains enhanced stability. Using live cell imaging and flow cytometry, we demonstrated that HBcAg VLPs functionalized with CPP exhibited up to 40% more efficient penetration into professional antigen-presenting cells (JAWS II) than HBcAg VLPs without CPP. This resulted in the increased presentation of integrated T cell epitopes by dendritic cells. In vivo, we detected significantly increased induction of SV40 Large T antigen-specific CTL in mice immunized with CPP-conjugated VLPs compared to unconjugated VLPs. In this study, we demonstrated that a stronger cellular immune response can be induced by CPP-functionalized HBcAg VLPs than with the unmodified HBcAg VLPs in vitro as well as in vivo. This discovery may have positive implications for future vaccine development where an enhanced cellular component of the immune response is desirable.

Posttranslational modifications are involved in the regulation of virtually all cellular processes, including immune response, nevertheless, they are also targets manipulated by invading pathogens. The first investigated example is protein citrullination which is an important posttranslational modification that acts on a multitude of processes like supervision of cell pluripotency and rheumatoid arthritis. Citrullination of targeted arginine residues is performed by the Peptidylarginine deiminase. Within the first published manuscript, being part of this thesis, it was possible to show the use of this posttranslational modification by the human pathogen Porphyromonas gingivalis to facilitate innate immune evasion at three distinct level. P. gingivalis was demonstrated to citrullinate proteins by Porphyromonas peptidylarginine deiminase resulting in diminished phagocytosis and subsequent killing by neutrophils. Furthermore, it was shown that citrullination of histone H3 enables P. gingivalis to survive in neutrophil extracellular traps and incapacitate the lysozyme-derived peptide LP9. The second investigated posttranslational modification is ubiquitination and its role in respiratory tract infections. Ubiquitination is the covalent attachment of a small protein that consisting of only 76 amino acids to the ε-amino group of lysine residues to posttranslational modify proteins. Acute infections of the lower respiratory tract such as viral and bacterial co-infections are among the most prevalent reasons of fatal casualties worldwide. Therefore, the interactions between host and pathogens resulting in the impairment of the hosts immune response and immune evasion of the pathogens, need to be elucidated. To get new insights in the infection driven changes in protein polyubiquitination and alterations in the abundance of ubiquitin E3 ligases involved in ubiquitination, cellular proteomes were monitored in detail by high resolution mass spectrometry. Therefore, the epithelial cell lines 16HBE14o- (Manuscript II) and A549 (Manuscript III) were co-infected with influenza A virus H1N1 and Streptococcus pyogenes or Staphylococcus aureus or with influenza A virus H1N1 and Streptococcus pneumoniae, respectively. Here, it could be shown in 16HBE14o- cells that co-infection of epithelial cells is not characterized by decreased cell survival and that observable effects on the proteome and ubiquitinome are mostly additive rather than synergistic. S. pyogenes infection affected the mitochondrial function, cell-cell adhesion, endocytosis and actin organization. Viral infection affected mRNA processing and Rho signaling. Viral and bacterial co-infection was detected to affect processes that were already affected by both of the corresponding single infections. No further pathways were strongly affected by the co-infection. A similar result has been observed in A549 cells co-infected IAV and S. pneumoniae. Overrepresented gene ontology terms depict the sum of those observed in the viral and bacterial single infection. Moreover, no significant change in cell survival upon co-infection compared to single bacterial infection was noticed for A549 cells either. This led to the suggestion that co-infection of investigated epithelial cells under examined conditions possesses additive rather than synergistic effect and thus, may not worsen the outcome of the infection within the studied conditions. Infections in other systems, may provide varying results and thus should be examined in future studies.

Das Afrikanische Schweinepestvirus (ASPV) ist ein wirtschaftlich wichtiger und in Haus- und Wildschweinen Hämorrhagie mit hoher Sterblichkeitsrate verursachender viraler Erreger. 1921 erstmals in Kenia beschrieben, breitete sich die ASP seit 2007 auch über den Kaukasus, ins Baltikum (2014), weiter in europäische und asiatische Länder und seit 2020 in Deutschland aus. Trotz der hohen genetischen Stabilität des Afrikanischen Schweinepestvirus (ASPV) wurden Genomvarianten identifiziert, bei denen Unterschiede in der Genexpression von Multigenfamilien (MGF) dominieren. Letztlich divergieren ASPV-Stämme in ihrer Virulenz und verursachen akut-letale bis chronische Verläufe im Schwein. Aufgrund der enormen Komplexität des Virus und seiner vielfältigen Immunevasionsstrategien sind viele Mechanismen der Virus-Wirts-Interaktion, die zur Immunpathogenese beitragen, nicht ausreichend verstanden und erschweren somit die Impfstoffentwicklung. Dabei können virale Subversionsmechanismen der Wirtszelle die antivirale Immunantwort modulieren und stehen deshalb im Fokus dieser Arbeit. Zur Charakterisierung und mechanistischen Aufklärung dieser ASPV-spezifischen Immunsubversionsmechanismen wurden primäre porzine Monozyten von Hausschweinen mit hochvirulentem (Armenia) und natürlich-attenuiertem (Estonia) ASPV infiziert. Die Resultate ergaben sowohl stammunabhängige als auch -abhängige Unterschiede in der Regulation myeloider Oberflächenmarker infizierter Monozyten. Insbesondere beobachteten wir eine stammunabhängige Suppression des Phagozytose-regulierenden CD172a und eine stammabhängige Regulation von porzinem MHC I (SLA I). Weitere Experimente zur Untersuchung der zugrundeliegenden Mechanismen ergaben, dass zwar beide Stämme die Oberflächenexpression von CD172 unterdrücken, jedoch nur Armenia-, im Gegensatz zu Estonia-infizierten Monozyten, eine reduzierte Recyclingrate sowie eine Abspaltung (Shedding) von CD172a von der Zelloberfläche zeigten. Dies lässt vermuten, dass die Virus-vermittelte Suppression von CD172a der beiden ASPV-Stämme auf unterschiedlichen Subversionsmechanismen beruht. Reinfektionsexperimente und molekularbiologische Untersuchungen belegten zudem, dass das abgespaltene Oberflächen-CD172a der Armenia-infizierten Monozyten mit einer gesteigerten Infektionsrate einhergeht, dies ist wahrscheinlich das Ergebnis (entweder direkt oder indirekt) einer Komplexbildung zwischen dem virulenten Armenia-Virus und löslichem CD172a. Im Gegensatz dazu resultierte die Infektion von Monozyten mit Armenia, jedoch nicht mit Estonia, in einem deutlichen Oberflächenverlust von porzinem SLA I, welches für die Antigenpräsentation gegenüber CD8+ T-Zellen essentiell ist. Weitere Versuche zeigten einen Reifungsdefekt von SLA I, der mit dem Abbau funktioneller ER-Strukturen und der Induktion von ER-Stress in Armenia-infizierten Monozyten in Zusammenhang stand. Gleichzeitig wurde eine deutlich reduzierte Überlebensfähigkeit Armenia-infizierter Monozyten beobachtet, die mit einem Verlust mitochondrialer Funktionen und der Bildung von Aggresomen aus fehlgefalteten Proteinen im Zytoplasma einherging. Vertiefende Analysen dazu zeigten einen Caspase-3 aktivierten Zelltodmechanismus und ein infektionsbedingtes, progressives Abschalten der Proteintranslation in Armenia-infizierten Zellen. Um einen möglichen Zusammenhang zwischen den beobachteten Subversionsmechanismen und der Expression bestimmter viraler MGF-Gene zu finden, wurden weitere ASPV-Stämme in die Untersuchungen zur CD172a- und SLA I-Oberflächenexpression einbezogen. Ähnlich wie Armenia zeigte sich auch für die Stämme NHV und OURT88/3 eine deutliche Reduktion der SLA I-Oberflächenlevel, auch wenn diese in vivo gering-virulent sind. Andererseits zeigte das hochvirulente Benin97/1-Isolat im Gegensatz zu Armenia keine SLA I-Subversion, sondern ähnlich wie nach Estonia-Infektion kaum veränderte SLA I-Level, was vermuten lässt, dass der SLA I Subversionsmechanismus nicht alleinig den Virulenzgrad der ASPV-Stämme bestimmt. Ein direkter Genomvergleich identifizierte verschiedene Mitglieder der MGF110- und MGF505-Gene als möglicherweise beteiligte virale Genkandidaten. Im Gegensatz hierzu ergaben sich keine detektierbaren Unterschiede bei den Analysen zur Oberflächensuppression von CD172a innerhalb der verwendeten Isolate, wie bereits bei Armenia und Estonia Infektion beobachtet. Interessanterweise beobachteten wir dabei das Vorhandensein von MGF110-14 als eine genomische Gemeinsamkeit, die für die generelle Oberflächenreduktion von CD172a, zusätzlich zu anderen Genen, die ein Shedding und die Armenia-spezifische Interaktion bestimmen könnten, verantwortlich sein könnte. Insgesamt zeigen die Resultate dieser Arbeit erstmals, dass das virulente ASPV Armenia, anders als das attenuierte ASPV Estonia, einen ausgeprägten Funktions- und Vitalitätsverlust in seinen primären Zielzellen (z. B. Monozyten) bewirkt. Die gesteigerte Infektiosität, Induktion von zellulärem Stress und Beeinträchtigung der SLA I-vermittelten Antigenpräsentation werden in infizierten Schweinen eine entscheidende Rolle in der Virus-Verbreitung und der Immunevasion spielen. Zusammenfassend lässt sich sagen, dass die Befunde dieser Arbeit neue und vertiefte Einblicke in die zellulären Mechanismen der SLA I- und CD172a-Subversion im Zusammenhang mit der Immunevasion durch hoch-virulentes ASPV Armenia und attenuiertes ASPV Estonia gibt und zudem wichtig für das bessere Verständnis der ASP-Immunpathogenese sind.

Methane (CH4) is a potent greenhouse gas with rising atmospheric concentrations. Microorganisms are essential players in the global methane cycle. In fact, the largest part of methane emissions derives from microbial production by methanogenic Archaea (methanogens). Microorganisms do not only produce methane: methanotrophs can also oxidize the methane produced by methanogens. In addition, soil methanotrophs are the only biological methane sink, oxidizing up to 30-40 Tg of this potent greenhouse gas per year worldwide. However, intensified management of grasslands and forests may reduce the methane sink capacity of soils. In general, the interaction of methanogens and methanotrophs determines whether a soil is a source or a sink for methane. It is, therefore, crucial to understand the microbial part of the methane cycle and which factors influence the abundance and activity of methane-cycling microbes. However, capturing the soil microbiome's abundances, activity, and identity is challenging. There are numerous target molecules and myriad methods, each with certain limitations. Linking microbial markers to methane fluxes is therefore challenging. This thesis aimed to understand how methane-cycling microbes in the soil are related to soil methane fluxes and how soil characteristics and human activity influence them. The first publication investigated the biotic and abiotic drivers of the atmospheric methane sink of soils. It assessed the influence of grassland land-use intensity (150 sites) and forest management type (149 sites) on potential atmospheric methane oxidation rates (PMORs) and the abundance and diversity of CH4-oxidizing bacteria (MOB) with qPCR in topsoils of three temperate regions in Germany. PMORs measured in microcosms under defined conditions were approximately twice as high in forest than in grassland soils. High land-use intensity of grasslands negatively affected PMORs (−40%) in almost all regions. Among the different aspects of land-use intensity, fertilization had the most adverse effect reducing PMORs by 20%. In contrast, forest management did not affect PMORs in forest soils. Upland soil cluster (USC)α was the dominant group of MOBs in the forests. In contrast, USCγ was absent in more than half of the forest soils but present in almost all grassland soils. USCα abundance had a direct positive effect on PMOR in forests, while in grasslands, USCα and USCγ abundance affected PMOR positively with a more pronounced contribution of USCγ than USCα. In the second publication, we used quantitative metatranscriptomics to link methane-cycling microbiomes to net surface methane fluxes throughout a year in two grassland soils. Methane fluxes were highly dynamic: both soils were net methane sources in autumn and winter and net methane sinks in spring and summer. Correspondingly, methanogen mRNA abundances per gram soil correlated well with methane fluxes. Methanotroph to methanogen mRNA ratios were higher in spring and summer when the soils acted as net methane sinks. Furthermore, methane uptake was associated with an increased proportion of USCα and γ pmoA and pmoA2 transcripts. High methanotroph to methanogen ratios would indicate methane sink properties. Our study links the seasonal transcriptional dynamics of methane-cycling soil microbiomes for the first time to gas fluxes in situ. It suggests mRNA transcript abundances as promising indicators of dynamic ecosystem-level processes. We conclude that reduction in grassland land-use intensity and afforestation can potentially increase the methane sink function of soils and that different parameters determine the microbial methane sink in forest and grassland soils. Furthermore, this thesis suggests mRNA transcript abundances as promising indicators of dynamic ecosystem-level processes. Methanogen transcript abundance may be used as a proxy for changes in net surface methane emissions from grassland soils.

Coding constraints imposed by the very small genome sizes of negative-strand RNA viruses (NSVs) have led to the development of numerous strategies that increase viral protein diversity, enabling the virus to both establish a productive viral replication cycle and effectively control the host antiviral response. Arenaviruses are no exception to this, and previous findings have demonstrated that the nucleoprotein (NP) of the highly pathogenic Junín virus (JUNV) exists as three additional N-terminally truncated isoforms of 53 kD (NP53kD), 47 kD (NP47kD), and 40 kD (NP40kD). The two smaller isoforms (i.e. NP47kD and NP40kD) have been characterized as products of caspase cleavage, which appears to serve a decoy function to inhibit apoptosis induction. However, whether they have additional functions in the viral replication cycle remains unknown. Further, the origin and function of NP53kD has not yet been described. In order to first identify the mechanism responsible for production of the NP53kD variant, a possible role of additional caspase cleavage sites was first excluded using a site mutagenesis approach. Subsequently, alanine mutagenesis was then used to identify a region responsible for NP53kD production. As a result, three methionine residues were identified within the characterized sequence segment of NP, linking the production of NP53kD to an alternative in-frame translation initiation. Further site-directed mutagenesis of the previously identified putative in-frame methionine codons (i.e. M78, M80 and M100) finally led to the identification of translation initiation at M80 as being predominantly responsible for the production of NP53kD. Once the identity of all three NP isoforms was known, it was then of further interest to more deeply characterize their functional roles. Consistent with the N-terminal domain containing RNA binding and homotrimerization motifs that are relevant for the viral RNA synthesis process, it could be demonstrated that all three truncated NP isoforms lost the ability to support viral RNA synthesis in a minigenome assay. However, they also did not interfere with viral RNA synthesis by full-length NP, nor did they affect the ability of the matrix protein Z to inhibit viral RNA synthesis. Moreover, it was observed that loss of the oligomerization motifs in the N-terminus also affected the subcellular localization of all three NP isoforms, which were no longer localized in discrete perinuclear inclusion bodies, but rather showed a diffuse distribution throughout the cytoplasm, with the smallest isoform NP40kD also being able to enter the nucleus. Surprisingly, the 3'-5' exonuclease function of NP, which is associated with the C-terminal domain and plays a role in inhibiting interferon induction by digestion of double-stranded RNAs, was found to be retained only by the NP40kD isoform, despite that all three isoforms retained the associated domain. Finally, previous studies using transfected NP and chemical induction of apoptosis have suggested that cleavage of NP at the caspase motifs responsible for generating NP47kD and NP40kD plays a role in controlling activation of the apoptosis pathway. Therefore, to further characterize the connection between the generation of NP isoforms and the regulation of apoptosis in a viral context, recombinant JUNVs deficient in the respective isoforms were generated. Unlike infections with wild-type JUNV, mutations of the caspase cleavage sites resulted in the induction of caspases activation. Surprisingly, however, this was also the case for mutation of the alternate start codon responsible for NP53kD generation. Taken together, the data from this study suggest a model whereby JUNV generates a pool of smaller NP isoforms with a predominantly cytoplasmic distribution. As a result of this altered localization, NP53kD appears to be able to serve as the substrate for further generation of NP47kD and NP40kD by caspase cleavage. Not only does this cleavage inhibit apoptosis induction during JUNV infection, it also results in a cytoplasmic isoform of NP that retains strong 3'-5' exonuclease activity (i.e. NP40kD) and thus may play an important role in preventing viral double-stranded RNA accumulation in the cytoplasm, where it can lead to activation of IFN signaling. Overall, such results emphasize the relevance of alternative protein isoforms in virus biology, and particularly in regulation of the host response to infection.