Search

Recoding viral genomes by introducing numerous synonymous nucleotide substitutions that create suboptimal codon pairs provides new live-attenuated vaccine candidates. Because recoding typically involves a large number of nucleotide substitutions, the risk of de-attenuation is presumed to be low. However, this has not been thoroughly studied. We previously generated human respiratory syncytial virus (RSV) in which the NS1, NS2, N, P, M and SH ORFs were codon-pair deoptimized (CPD) by 695 synonymous nucleotide changes (Min A virus). Min A exhibited a global reduction in transcription and protein synthesis, was restricted for replication in vitro and in vivo, and exhibited moderate temperature sensitivity. Here, we show that under selective pressure by serial passage at progressively increasing temperatures, Min A regained replication fitness and lost its temperature sensitivity. Whole-genome deep sequencing identified numerous missense mutations in several genes, in particular ones accumulating between codons 25 and 34 of the phosphoprotein (P), a polymerase cofactor and chaperone. When re-introduced into Min A, these P mutations restored viral transcription to wt level, resulting in increased protein expression and RNA replication. Molecular dynamic simulations suggested that these P mutations increased the flexibility of the N-terminal domain of P, which might facilitate its interaction with the nucleoprotein N, and increase the functional efficiency of the RSV transcription/replication complex. Finally, we evaluated the effect of the P mutations on Min A replication and immunogenicity in hamsters. Mutation P[F28V] paradoxically reduced Min A replication but not its immunogenicity. The further addition of one missense mutation each in M and L generated a version of Min A with increased genetic stability. Thus, this study provides further insight into the adaptability of large-scale recoded RNA viruses under selective pressure and identified an improved CPD RSV vaccine candidate., Author summary Synonymous recoding of viral genomes by codon-pair deoptimization (CPD) generates live-attenuated vaccines presumed to be genetically stable due to the high number of nucleotide substitutions. However, their actual genetic stability under selective pressure was largely unknown. In a recoded human respiratory syncytial virus (RSV) mutant called Min A, six of 11 ORFs were CPD, reducing protein expression and inducing moderate temperature sensitivity and attenuation. When passaged in vitro under selective pressure, Min A lost its temperature-sensitive phenotype and regained fitness by the acquisition of numerous mutations, in particular missense mutations in the viral phosphoprotein (P), a polymerase cofactor and a chaperone for soluble nucleoprotein. These P mutations increased RSV gene transcription globally, thereby increasing RSV protein expression, RNA replication, and virus particle production. Thus, the P mutations increased the efficiency of the RSV transcription/replication complex, compensating for the reduced protein expression due to CPD. In addition, introduction of the P mutations into Min A generated a live-attenuated vaccine candidate with increased genetic stability. Surprisingly, this vaccine candidate exhibited increased attenuation and, paradoxically, exhibited increased immunogenicity per plaque-forming unit in hamsters. This study provides insights into the adaptability of large-scale recoded RNA viruses and identified an improved CPD RSV vaccine candidate.

Tendons and ligaments are critical components of the musculoskeletal system, yet the pathways specifying this lineage remain poorly defined. Through a screen of known bioactive chemicals in zebrafish, we identified a new pathway regulating tendon cell induction. We established that statin, through inhibition of the mevalonate pathway, causes an expansion of the tendon progenitor population. Co-expression and live imaging studies indicate that the expansion does not involve an increase in cell proliferation, but rather results from re-specification of cells from the neural crest-derived sox10+/sox9a+ skeletal lineage. The effect on tendon cell expansion is specific to the geranylgeranylation branch of the mevalonate pathway and mediated by inhibition of Rac activity. This work establishes a novel role for the mevalonate pathway and Rac activity in regulating specification of the tendon lineage.

Integrated development of diverse tissues gives rise to a functional, mobile vertebrate musculoskeletal system. However, the genetics and cellular interactions that drive the integration of muscle, tendon, and skeleton are poorly understood. In the vertebrate head, neural crest cells, from which cranial tendons derive, pattern developing muscles just as tendons have been shown to in limb and trunk tissue, yet the mechanisms of this patterning are unknown. From a forward genetic screen, we determined that cyp26b1 is critical for musculoskeletal integration in the ventral pharyngeal arches, particularly in the mandibulohyoid junction where first and second arch muscles interconnect. Using time-lapse confocal analyses, we detail musculoskeletal integration in wild-type and cyp26b1 mutant zebrafish. In wild-type fish, tenoblasts are present in apposition to elongating muscles and condense in discrete muscle attachment sites. In the absence of cyp26b1, tenoblasts are generated in normal numbers but fail to condense into nascent tendons within the ventral arches and, subsequently, muscles project into ectopic locales. These ectopic muscle fibers eventually associate with ectopic tendon marker expression. Genetic mosaic analysis demonstrates that neural crest cells require Cyp26b1 function for proper musculoskeletal development. Using an inhibitor, we find that Cyp26 function is required in a short time window that overlaps the dynamic window of tenoblast condensation. However, cyp26b1 expression is largely restricted to regions between tenoblast condensations during this time. Our results suggest that degradation of RA by this previously undescribed population of neural crest cells is critical to promote condensation of adjacent scxa-expressing tenoblasts and that these condensations are subsequently required for proper musculoskeletal integration., Author summary Mobility requires that muscles form appropriate attachments via tendon. The genes regulating this attachment are poorly understood. This is especially true in the head where mesoderm and neural crest cells generate muscle and tendon, respectively. We show that the gene cyp26b1 is critical for muscle tendon attachment in a specific region of the developing head. The movement of tendon generating cells into mature tendons requires cyp26b1. Loss of cyp26b1 disrupts these movements and muscles subsequently extend into inappropriate areas. Cyp26b1 appears to be required in neural crest cells but not the tendon generating cells. Collectively, our work provides insight into the genetic and cellular mechanisms that attach muscle and tendon.

Despite the importance of tendons and ligaments for transmitting movement and providing stability to the musculoskeletal system, their development is considerably less well understood than that of the tissues they serve to connect. Zebrafish have been widely used to address questions in muscle and skeletal development, yet few studies describe their tendon and ligament tissues. We have analyzed in zebrafish the expression of several genes known to be enriched in mammalian tendons and ligaments, including scleraxis (scx), collagen 1a2 (col1a2) and tenomodulin (tnmd), or in the tendon-like myosepta of the zebrafish (xirp2a). Co-expression studies with muscle and cartilage markers demonstrate the presence of scxa, col1a2 and tnmd at sites between the developing muscle and cartilage, and xirp2a at the myotendinous junctions. We determined that the zebrafish craniofacial tendon and ligament progenitors are neural crest derived, as in mammals. Cranial and fin tendon progenitors can be induced in the absence of differentiated muscle or cartilage, although neighboring muscle and cartilage are required for tendon cell maintenance and organization, respectively. By contrast, myoseptal scxa expression requires muscle for its initiation. Together, these data suggest a conserved role for muscle in tendon development. Based on the similarities in gene expression, morphology, collagen ultrastructural arrangement and developmental regulation with that of mammalian tendons, we conclude that the zebrafish tendon populations are homologous to their force-transmitting counterparts in higher vertebrates. Within this context, the zebrafish model can be used to provide new avenues for studying tendon biology in a vertebrate genetic system.