Biology subjects

de Wit, E.

Publications and source records attributed to de Wit, E..

At least 37 records · Page 2Linked to original sources

bioRxiv2022

Identifying cross-lineage dependencies of cell-type specific regulators in gastruloids

Correct gene expression levels in space and time are crucial for normal development. Advances in genomics enable the inference of gene regulatory programs that are active during development. However, this approach cannot capture the complex multicellular interactions that occur during embryogenesis. Compared to model organisms such as fruit flies and zebrafish, the growth of mammalian embryos in utero further complicates the analysis of cell-cell communication during development. However, in vitro models of mammalian development such as gastruloids can overcome this limitation. Using time-resolved single-cell chromatin accessibility analysis, we have delineated the regulatory landscape during gastruloid development and thereby identified the critical drivers of developmental transitions. We observed that gastruloids develop from pluripotent cells driven by the transcription factor (TF) dimer OCT4-SOX2 and differentiate along two main branches. A mesoderm branch characterized by the TF MSGN1 and a spinal cord branch characterized by CDX1, 2, 4 (CDX). Consistent with our lineage reconstruction, {Delta}CDX gastruloids fail to form spinal cord. Conversely, Msgn1 ablation inhibits the development of paraxial mesoderm, as expected. However, this also abolished spinal cord cells, which is surprising given that MSGN1 is not associated with differentiation along this branch. Therefore, formation of paraxial mesoderm is required for spinal cord development. To validate this, we generated chimeric gastruloids using {Delta}MSGN1 and wildtype cells, which formed both spinal cord and paraxial mesoderm. Strikingly, {Delta}MSGN1 cells specifically contributed to spinal cord, suggesting that cell-cell interactions between paraxial mesoderm and spinal cord are necessary for the formation of the latter. Our work has important implications for the study of cell-cell communication in development and how gene regulatory programs are functionally executed to form complex multicellular developmental structures.

systems biology↗

bioRxiv2022

SARS-CoV-2 Omicron BA.1 and BA.2 are attenuated in rhesus macaques as compared to Delta

Since the emergence of SARS-CoV-2, five different variants of concern (VOCs) have been identified: Alpha, Beta, Gamma, Delta, and Omicron. Due to confounding factors in the human population, such as pre-existing immunity, comparing severity of disease caused by different VOCs is challenging. Here, we investigate disease progression in the rhesus macaque model upon inoculation with the Delta, Omicron BA.1, and Omicron BA.2 VOCs. Disease severity in rhesus macaques inoculated with Omicron BA.1 or BA.2 was lower than those inoculated with Delta and resulted in significantly lower viral loads in nasal swabs, bronchial cytology brush samples, and lung tissue in rhesus macaques. Cytokines and chemokines were upregulated in nasosorption samples of Delta animals compared to Omicron BA.1 and BA.2 animals. Overall, these data suggests that in rhesus macaques, Omicron replicates to lower levels than the Delta VOC, resulting in reduced clinical disease.

microbiology↗

bioRxiv2022

Pioneer activity distinguishes activating from non-activating pluripotency transcription factor binding sites

Genome-wide transcriptional activity involves the binding of many transcription factors to thousands of sites in the genome. Determining which sites are directly driving transcription remains a challenge. Here we use acute protein depletion of the pioneer transcription factors OCT4 and SOX2 to establish their functionality in maintaining chromatin accessibility. We show that thousands of accessible sites are lost within an hour of protein depletion, indicating rapid turnover of these sites in the absence of pioneer factors. To understand the relationship with transcription we performed nascent transcription analysis and found that open chromatin sites that are maintained by SOX2 are highly predictive of gene expression, in contrast to SOX2 binding sites that do not maintain accessibility. We use CRISPR-Cas9 genome editing in the Klf2 locus to functionally validate a predicted regulatory element. We conclude that the regulatory activity of SOX2 is exerted largely at sites where it maintains accessibility and that other binding sites are largely dispensable for gene regulation.

molecular biology↗

bioRxiv2022

Mild SARS-CoV-2 infection in rhesus macaques is associated with viral control prior to antigen-specific T cell responses in tissues

SARS-CoV-2 primarily replicates in mucosal sites, and more information is needed about immune responses in infected tissues. We used rhesus macaques to model protective primary immune responses in tissues during mild COVID-19. Viral RNA levels were highest on days 1-2 post-infection and fell precipitously thereafter. 18F-fluorodeoxyglucose (FDG)-avid lung abnormalities and interferon (IFN)-activated myeloid cells in the bronchoalveolar lavage (BAL) were found on days [~]3-4. Virus-specific effector CD8 and CD4 T cells were detectable in the BAL and lung tissue on days [~]7-10, after viral RNA, lung inflammation, and IFN-activated myeloid cells had declined. Notably, SARS-CoV-2-specific T cells were not detectable in the nasal turbinates, salivary glands, and tonsils on day 10 post-infection. Thus, SARS-CoV-2 replication wanes in the lungs prior to T cell responses, and in the nasal and oral mucosa despite the apparent lack of Ag-specific T cells, suggesting that innate immunity efficiently restricts viral replication during mild COVID-19. ONE SENTENCE SUMMARYSARS-CoV-2 infection leads to mild, focal lung inflammation, and type I IFN activated myeloid cells that mostly resolve prior to the influx of virus-specific effector T cells or antibody responses in rhesus macaques.

immunology↗

bioRxiv2021

Distinct roles for CDK-Mediator in controlling Polycomb-dependent chromosomal interactions and priming genes for induction

Precise control of gene expression underpins normal development. This relies on mechanisms that enable communication between gene promoters and other regulatory elements. In embryonic stem cells (ESCs), the CDK-Mediator (CDK-MED) complex has been reported to physically link gene regulatory elements to enable gene expression and also prime genes for induction during differentiation. Here we discover that CDK-MED contributes little to 3D genome organisation in ESCs, but has a specific and essential role in controlling interactions between inactive gene regulatory elements bound by Polycomb repressive complexes (PRCs). These interactions are established by the canonical PRC1 (cPRC1) complex but rely on CDK-MED, which facilitates binding of cPRC1 to its target sites. Importantly, through separation of function experiments, we reveal that this collaboration between CDK-MED and cPRC1 in creating long-range interactions does not function to prime genes for induction during differentiation. Instead, we discover that priming relies on an interaction-independent mechanism whereby the CDK module supports core Mediator engagement with gene promoters to enable gene activation.

molecular biology↗

bioRxiv2021

CTCF and cohesin promote focal detachment of DNA from the nuclear lamina

Lamina associated domains (LADs) are large genomic regions that are positioned at the nuclear lamina (NL). It has remained largely unclear what drives the positioning and demarcation of LADs. Because the insulator protein CTCF is enriched at LAD borders, it was postulated that CTCF binding could position a subset of LAD boundaries, possibly through its function in stalling cohesin and hence preventing cohesin to invade into the LAD. To test this, we mapped genome - NL interactions in mouse embryonic stem cells after rapid depletion of CTCF and other perturbations of cohesin dynamics. CTCF and cohesin contribute to a sharp transition in NL interactions at LAD borders, whilst LADs are maintained after depletion of these proteins, also at borders marked by CTCF. CTCF and cohesin may thus reinforce LAD borders, but do not position these. CTCF binding sites within LADs are locally detached from the NL and enriched for accessible DNA and active histone modifications. Remarkably, even though NL positioning is strongly correlated with genome inactivity, this DNA remains accessible after the local detachment is lost following CTCF depletion. At a chromosomal scale, cohesin depletion and cohesin stabilization (depletion of the unloading factor WAPL) quantitatively affect NL interactions, indicative of perturbed chromosomal positioning in the nucleus. Finally, while H3K27me3 is locally enriched at CTCF-marked LAD borders, we find no evidence for an interplay between CTCF and H3K27me3 on NL interactions. Combined, these findings illustrate that CTCF and cohesin do not shape LAD patterns. Rather, these proteins mediate fine-tuning of NL interactions.

molecular biology↗

bioRxiv2021

Age-related differences in immune dynamics during SARS-CoV-2 infection in rhesus macaques

Advanced age is a key predictor of severe COVID-19. To gain insight into this relationship, particularly with respect to immune responses, we utilized the rhesus macaque model of SARS-CoV-2 infection. Two cohorts of eight older (16-23 years) and eight younger (3-5 years) rhesus macaques were inoculated with SARS-CoV-2. Animals were evaluated using viral RNA quantification, clinical observations, thoracic radiographs, single-cell transcriptomics, multiparameter flow cytometry, multiplex immunohistochemistry, cytokine detection, and lipidomics analysis at pre-defined timepoints in various tissues. Differences in clinical signs, pulmonary infiltrates, and virus replication dynamics were limited between age cohorts. Transcriptional signatures of inflammation-associated genes in cells isolated from bronchoalveolar lavage fluid at 3 dpi revealed efficient mounting of innate immune defenses in both younger and older animals. These findings suggested that age did not substantially skew major facets of acute disease in this model. However, age-specific divergence of immune responses emerged during the post-acute phase of infection (7-21 dpi). Older animals exhibited sustained local inflammatory innate responses while local effector T-cell responses were induced earlier in the younger animals. Circulating lipid mediator and cytokine levels highlighted increased repair-associated signals in the younger animals, in contrast to persistent pro-inflammatory responses in the older animals. In summary, despite similar disease outcomes, multi-omics profiling in SARS-CoV-2-infected rhesus macaques suggests that age may delay or impair the induction of anti-viral cellular immune responses and delay efficient return to immune homeostasis following acute infection.

immunology↗

bioRxiv2021

Rapid depletion of CTCF and cohesin proteins reveals dynamic features of chromosome architecture.

The interphase genome is mainly shaped by cohesin-mediated loop extrusion and cohesin-independent compartmentalization. Extrusion is a dynamic process of cohesin loading, loop extension and release. Cohesin release is mediated by WAPL. Loss of WAPL leads to the formation of longer loops and counters compartmentalization. The dynamics of these changes in chromosome organization have been unclear. We have used acute depletion of WAPL to show that within six hours cohesin accumulates at CTCF-bound loop anchors and extended loops are formed. When we deplete WAPL and CTCF simultaneously, new loops are formed between active genes. Surprisingly, active gene clustering is independent of cohesin. Stabilization of cohesin on chromatin leads to a decrease in compartmentalization, which is rapidly restored by depletion of cohesin. Our analyses show that loop extrusion counters compartmentalization and plays a central role in many aspects of chromosome organization. HIGHLIGHTSO_LICohesin accumulates at CTCF-mediated chromatin loop anchors following WAPL depletion. C_LIO_LIActively transcribed genes form long-range gene clusters independent of the cohesin complex. C_LIO_LIPlumes are a novel architectural feature of juxtaposed DNA formed by cohesin at open chromatin islands. C_LIO_LIChromosome compartmentalization can be uncoupled from nuclear lamina interactions. C_LI

molecular biology↗

bioRxiv2021

Unexpected gene activation following CRISPR-Cas9-mediated genome editing

The discovery of the Clustered Regularly-Interspaced Short Palindromic Repeats (CRISPR) and its development as a genome editing tool has revolutionized the field of molecular biology. In the DNA damage field, CRISPR has brought an alternative to induce endogenous double-strand breaks (DSB) at desired genomic locations and study the DNA damage response and its consequences. Many systems for sgRNA delivery have been reported in order to efficiency generate this DSB, including lentiviral vectors. However, some of the consequences of these systems are yet not well understood. Here we report that lentiviral-based sgRNA vectors can integrate into the endogenous genomic target location, leading to undesired activation of the target gene. By generating a DSB in the regulatory region of the ABCB1 gene using a lentiviral sgRNA vector, we can induce the formation of taxol-resistant colonies. We show that these colonies upregulated ABCB1 via integration of the EEF1A1 and the U6 promoters from the sgRNA vector. We believe that this is an unreported CRISPR/Cas9 artefact that researchers need to be aware of when using lentiviral vectors for genome editing.

genomics↗

bioRxiv2021

Subtle differences in the pathogenicity of SARS-CoV-2 variants of concern B.1.1.7 and B.1.351 in rhesus macaques

The emergence of several SARS-CoV-2 variants has caused global concerns about increased transmissibility, increased pathogenicity, and decreased efficacy of medical countermeasures. Animal models can be used to assess phenotypical changes in the absence of confounding factors that affect observed pathogenicity and transmissibility data in the human population. Here, we studied the pathogenicity of variants of concern (VOC) B.1.1.7 and B.1.351 in rhesus macaques and compared it to a recent clade B.1 SARS-CoV-2 isolate containing the D614G substitution in the spike protein. The B.1.1.7 VOC behaved similarly to the D614G with respect to clinical disease, virus shedding and virus replication in the respiratory tract. Inoculation with the B.1.351 isolate resulted in lower clinical scores in rhesus macaques that correlated with lower virus titers in the lungs, less severe histologic lung lesions and less viral antigen detected in the lungs. We observed differences in the local innate immune response to infection. In bronchoalveolar lavages, cytokines and chemokines were upregulated on day 4 in animals inoculated with D614G and B.1.1.7 but not in those inoculated with B.1.351. In nasal samples, we did not detect upregulation of cytokines and chemokines in D614G or B.1.351-inoculated animals. However, cytokines and chemokines were upregulated in the noses of B.1.1.7-inoculated animals. Taken together, our comparative pathogenicity study suggests that ongoing circulation under diverse evolutionary pressure favors transmissibility and immune evasion rather than an increase in intrinsic pathogenicity.

microbiology↗

bioRxiv2021

Prior aerosol infection with lineage A SARS-CoV-2 variant protects hamsters from disease, but not reinfection with B.1.351 SARS-CoV-2 variant

The circulation of SARS-CoV-2 has resulted in the emergence of variants of concern (VOCs). It is currently unclear whether previous infection with SARS-CoV-2 provides protection against reinfection with VOCs. Here, we show that low dose aerosol exposure to hCoV-19/human/USA/WA-CDC-WA1/2020 (WA1, lineage A), resulted in a productive mild infection. In contrast, low dose of SARS-CoV-2 via fomites did not result in productive infection in the majority of exposed hamsters and these animals remained non-seroconverted. After recovery, hamsters were re-exposed to hCoV-19/South African/KRISP-K005325/2020 (VOC B.1.351) via an intranasal challenge. Seroconverted rechallenged animals did not lose weight and shed virus for 3 days. They had little infectious virus and no pathology in the lungs. In contrast, shedding, weight loss and extensive pulmonary pathology caused by B.1.351 replication was observed in the non-seroconverted animals. The rechallenged seroconverted animals did not transmit virus to naive sentinels via direct contact transmission, in contrast to the non-seroconverted animals. Reinfection with B.1.351 triggered an anamnestic response that boosted not only neutralizing titers against lineage A, but also titers against B.1.351. Our results confirm that aerosol exposure is a more efficient infection route than fomite exposure. Furthermore, initial infection with SARS-CoV-2 lineage A does not prevent heterologous reinfection with B.1.351 but prevents disease and onward transmission. These data suggest that previous SARS-CoV-2 exposure induces partial protective immunity. The reinfection generated a broadly neutralizing humoral response capable of effectively neutralizing B.1.351 while maintaining its ability to neutralize the virus to which the initial response was directed against.

microbiology↗

bioRxiv2021

ChAdOx1 nCoV-19 (AZD1222) protects against SARS-CoV-2 B.1.351 and B.1.1.7

We investigated ChAdOx1 nCoV-19 (AZD1222) vaccine efficacy against SARS-CoV-2 variants of concern (VOCs) B.1.1.7 and B.1.351 in Syrian hamsters. We previously showed protection against SARS-CoV-2 disease and pneumonia in hamsters vaccinated with a single dose of ChAdOx1 nCoV-19. Here, we observed a 9.5-fold reduction of virus neutralizing antibody titer in vaccinated hamster sera against B.1.351 compared to B.1.1.7. Vaccinated hamsters challenged with B.1.1.7 or B.1.351 did not lose weight compared to control animals. In contrast to control animals, the lungs of vaccinated animals did not show any gross lesions. Minimal to no viral subgenomic RNA (sgRNA) and no infectious virus was detected in lungs of vaccinated animals. Histopathological evaluation showed extensive pulmonary pathology caused by B.1.1.7 or B.1.351 replication in the control animals, but none in the vaccinated animals. These data demonstrate the effectiveness of the ChAdOx1 nCoV-19 vaccine against clinical disease caused by B.1.1.7 or B.1.351 VOCs.

microbiology↗

bioRxiv2021

Perturbations in 3D genome organization can promote acquired drug resistance

Acquired drug resistance is a major problem in the treatment of cancer. hTERT-immortalized, untransformed RPE-1 (RPE) cells can acquire resistance to taxol by derepressing the ABCB1 gene, encoding for the multidrug transporter P-gP. Here we have investigated how the ABCB1 gene is derepressed. We show that activation of the ABCB1 gene is associated with reduced DNA methylation, reduced H3K9 trimethylation and increased H3K27 acetylation at the ABCB1 promoter. In addition, we find that the ABCB1 locus has moved away from the nuclear lamina in the taxol-resistant cells. This raises the question which of these alterations were causal to derepression. Directly modifying DNA methylation or H3K27 methylation had neither significant effect on ABCB1 expression, nor did it promote drug resistance. In contrast, the disruption of Lamin B Receptor (LBR), a component of the nuclear lamina involved in genome organization, did promote the acquisition of a taxol-resistant phenotype in a subset of cells. Using CRISPRa-mediated gene activation, we could further substantiate a model in which disruption of lamina association renders the ABCB1 gene permissive to derepression. Based on these data we propose a model in which nuclear lamina dissociation of a repressed gene allows for its activation, implying that deregulation of the 3D genome topology could play an important role in tumor evolution and the acquisition of drug resistance.

molecular biology↗

bioRxiv2021

Hi-C Analyses with GENOVA: a case study with cohesin variants

Conformation capture-approaches like Hi-C can elucidate chromosome structure at a genome-wide scale. Hi-C datasets are large and require specialised software. Here, we present GENOVA: a user-friendly software package to analyse and visualise conformation capture data. GENOVA is an R-package that includes the most common Hi-C analyses, such as compartment and insulation score analysis. It can create annotated heatmaps to visualise the contact frequency at a specific locus and aggregate Hi-C signal over user-specified genomic regions such as ChIP-seq data. Finally, our package supports output from the major mapping-pipelines. We demonstrate the capabilities of GENOVA by analysing Hi-C data from HAP1 cell lines in which the cohesin-subunits SA1 and SA2 were knocked out. We find that {Delta}SA1 cells gain intra-TAD interactions and increase compartmentalisation. {Delta}SA2 cells have longer loops and a less compartmentalised genome. These results suggest that cohesinSA1 forms longer loops, while cohesinSA2 plays a role in forming and maintaining intra-TAD interactions. Our data supports the model that the genome is provided structure in 3D by the counter-balancing of loop formation on one hand, and compartmentalization on the other hand. By differentially controlling loops, cohesinSA1 and cohesinSA2 therefore also affect nuclear compartmentalization. We show that GENOVA is an easy to use R-package, that allows researchers to explore Hi-C data in great detail.

genomics↗

bioRxiv2020

SARS-CoV-2 infection dynamics in lungs of African green monkeys

Detailed knowledge about the dynamics of SARS-CoV-2 infection is important for unraveling the viral and host factors that contribute to COVID-19 pathogenesis. Old-World nonhuman primates recapitulate mild-moderate COVID-19 cases, thereby serving as important pathogenesis models. We compared African green monkeys inoculated with SARS-CoV-2 or inactivated virus to study the dynamics of virus replication throughout the respiratory tract. RNA sequencing of single cells from the lungs and mediastinal lymph nodes allowed a high-resolution analysis of virus replication and host responses over time. Viral replication was mainly localized to the lower respiratory tract, with evidence of replication in the pneumocytes. Macrophages were found to play a role in initiating a pro-inflammatory state in the lungs, while also interacting with infected pneumocytes. Our dataset provides a detailed view of changes in host and virus replication dynamics over the course of mild COVID-19 and serves as a valuable resource to identify therapeutic targets.

microbiology↗

bioRxiv2020

K18-hACE2 mice develop respiratory disease resembling severe COVID-19

SARS-CoV-2 emerged in late 2019 and resulted in the ongoing COVID-19 pandemic. Several animal models have been rapidly developed that recapitulate the asymptomatic to moderate disease spectrum. Now, there is a direct need for additional small animal models to study the pathogenesis of severe COVID-19 and for fast-tracked medical countermeasure development. Here, we show that transgenic mice expressing the human SARS-CoV-2 receptor (angiotensin-converting enzyme 2 [hACE2]) under a cytokeratin 18 promoter (K18) are susceptible to SARS-CoV-2 and that infection resulted in a dose-dependent lethal disease course. After inoculation with either 104 TCID50 or 105 TCID50, the SARS-CoV-2 infection resulted in rapid weight loss in both groups and uniform lethality in the 105 TCID50 group. High levels of viral RNA shedding were observed from the upper and lower respiratory tract and intermittent shedding was observed from the intestinal tract. Inoculation with SARS-CoV-2 resulted in upper and lower respiratory tract infection with high infectious virus titers in nasal turbinates, trachea and lungs. The observed interstitial pneumonia and pulmonary pathology, with SARS-CoV-2 replication evident in pneumocytes, were similar to that reported in severe cases of COVID-19. SARS-CoV-2 infection resulted in macrophage and lymphocyte infiltration in the lungs and upregulation of Th1 and proinflammatory cytokines/chemokines. Extrapulmonary replication of SARS-CoV-2 was observed in the cerebral cortex and hippocampus of several animals at 7 DPI but not at 3 DPI. The rapid inflammatory response and observed pathology bears resemblance to COVID-19. Taken together, this suggests that this mouse model can be useful for studies of pathogenesis and medical countermeasure development. Authors SummaryThe disease manifestation of COVID-19 in humans range from asymptomatic to severe. While several mild to moderate disease models have been developed, there is still a need for animal models that recapitulate the severe and fatal progression observed in a subset of patients. Here, we show that humanized transgenic mice developed dose-dependent disease when inoculated with SARS-CoV-2, the etiological agent of COVID-19. The mice developed upper and lower respiratory tract infection, with virus replication also in the brain after day 3 post inoculation. The pathological and immunological diseases manifestation observed in these mice bears resemblance to human COVID-19, suggesting increased usefulness of this model for elucidating COVID-19 pathogenesis further and testing of countermeasures, both of which are urgently needed.

microbiology↗

bioRxiv2020

Hydroxychloroquine Proves Ineffective in Hamsters and Macaques Infected with SARS-CoV-2

We remain largely without effective prophylactic/therapeutic interventions for COVID-19. Although many human clinical trials are ongoing, there remains a deficiency of supportive preclinical drug efficacy studies. Here we assessed the prophylactic/therapeutic efficacy of hydroxychloroquine (HCQ), a drug of interest for COVID-19 management, in two animal models. When used for prophylaxis or treatment neither the standard human malaria dose (6.5 mg/kg) nor a high dose (50 mg/kg) of HCQ had any beneficial effect on clinical disease or SARS-CoV-2 kinetics (replication/shedding) in the Syrian hamster disease model. Similarly, HCQ prophylaxis/treatment (6.5 mg/kg) did not significantly benefit clinical outcome nor reduce SARS-CoV-2 replication/shedding in the upper and lower respiratory tract in the rhesus macaque disease model. In conclusion, our preclinical animal studies do not support the use of HCQ in prophylaxis/treatment of COVID-19.One Sentence Summary Hydroxychloroquine prophylaxis/treatment showed no beneficial effect in SARS-CoV-2 hamster and macaque disease models.Competing Interest StatementThe authors have declared no competing interest.View Full Text

pharmacology and toxicology↗

bioRxiv2020

ChAdOx1 nCoV-19 vaccination prevents SARS-CoV-2 pneumonia in rhesus macaques

Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) emerged in December 20191,2 and is responsible for the COVID-19 pandemic3. Vaccines are an essential countermeasure urgently needed to control the pandemic4. Here, we show that the adenovirus-vectored vaccine ChAdOx1 nCoV-19, encoding the spike protein of SARS-CoV-2, is immunogenic in mice, eliciting a robust humoral and cell-mediated response. This response was not Th2 dominated, as demonstrated by IgG subclass and cytokine expression profiling. A single vaccination with ChAdOx1 nCoV-19 induced a humoral and cellular immune response in rhesus macaques. We observed a significantly reduced viral load in bronchoalveolar lavage fluid and respiratory tract tissue of vaccinated animals challenged with SARS-CoV-2 compared with control animals, and no pneumonia was observed in vaccinated rhesus macaques. Importantly, no evidence of immune-enhanced disease following viral challenge in vaccinated animals was observed. ChAdOx1 nCoV-19 is currently under investigation in a phase I clinical trial. Safety, immunogenicity and efficacy against symptomatic PCR-positive COVID-19 disease will now be assessed in randomised controlled human clinical trials.

microbiology↗