Not available

Striking rates of environmental changes combined with increased demand make it essential to develop a better understanding of global freshwater resources. Seasonal hypoxia in the central basin of Lake Erie is the result of thermal stratification and lake morphology. Limnetic physics can, however, only explain part of Lake Erie‟s “behavior”: the activity(s) of the ecosystem‟s biological members can be equally important. The goal of this study was to identify picocyanobacterial community members in the central basin of Lake Erie during summer stratification and the winter season to see how they may vary with season. Identification of microbial communities under the present environmental conditions establishes a relationship (i.e. baseline) from which changes can be seen over time. Seasonal variations in cyanobacterial communities can also offer insight into the biogeochemistry of Lake Erie. Information gained from the microbial ecology of Lake Erie can be applied to other lake systems.

Striking rates of environmental changes combined with increased demand make it essential to develop a better understanding of global freshwater resources. Seasonal hypoxia in the central basin of Lake Erie is the result of thermal stratification and lake morphology. Limnetic physics can, however, only explain part of Lake Erie‟s “behavior”: the activity(s) of the ecosystem‟s biological members can be equally important. The goal of this study was to identify picocyanobacterial community members in the central basin of Lake Erie during summer stratification and the winter season to see how they may vary with season. Identification of microbial communities under the present environmental conditions establishes a relationship (i.e. baseline) from which changes can be seen over time. Seasonal variations in cyanobacterial communities can also offer insight into the biogeochemistry of Lake Erie. Information gained from the microbial ecology of Lake Erie can be applied to other lake systems.

Sindbis virus (SINV) is a member of the Alphavirus genus and belongs to the family Togaviridae. The virus has a positive sense RNA genome of 11700 bases which encodes for both structural and non structural proteins. Infections are frequently diagnosed based on clinical, epidemiological and laboratory criteria. Laboratory confirmation is essential as SINV infections must be distinguished from various conditions that share similar clinical manifestations. The most frequently used methods for identification are haemagglutination inhibition, enzyme-linked immunosorbent assay, plaque reduction neutralization tests as well as conventional in-vitro neutralization assays. Serological assays for the detection of SINV are not readily available commercially and due to the non-specific symptoms caused by SINV infection the number of infections per annum may be under diagnosed. The purpose of this study was to develop serological assays such as ELISA and a novel neutralization assay that could be used in serological surveys for the detection of IgG antibodies against SINV. Furthermore to develop assays that could be used to determine the level of viral replication in mammalian cells for characterizing infection in mammalian cells as well as investigate the influence of interferon on viral replication and look for evidence of apoptosis caused by SINV infection. An in house ELISA was developed and used to screen 146 sera for IgG antibodies against SINV. The in-vitro neutralization assay is the gold standard for serology and 43 samples in total were tested in both the ELISA and the in-vitro neutralization assay. Analysis and comparison of the results obtained using the in-house ELISA and the neutralization assay indicated that the sensitivity of the ELISA was 68.9% and the specificity of the in house ELISA was 78.57 - 85.71% depending on the use of the percentage positive or optical density values to differentiate positive and negative samples. A forward and reverse primer for the amplification of a conserved 181bp region of the nsp2 gene encoding the nsp2 protein of SINV were designed along with a TaqMan hydrolysis probe to be used in a real time quantitative TaqMan PCR. The infection of mammalian cells, human macrophages and HeLa cells, was determined by measuring viral loads with a real time quantitative TaqMan RT-PCR. Two strains of SINV were used in attempts to infect macrophages, a strain from Egypt and a strain from South Africa. Small increases in viral load suggested possible low levels of viral replication but were considered insufficient to warrant further investigation and insufficient to investigate occurrence of antibody dependent enhancement of disease in macrophages. The mechanism possibly interfering with replication of virus in the human macrophages was investigated. Supernatant fluid samples from macrophage infections were tested for the release of interferon gamma which could inhibit viral replication. There were nine to fifteen fold differences in the concentration of interferon gamma detected in the supernatant fluid at baseline and 24h after infection. HeLa cells were treated with similar concentrations of human interferon gamma at different time intervals. Pretreatment and concurrent treatment with infection showed reduced levels of viral load compared with no treatment or delay in treatment. Hence the suggestion that interferon could have played a role in inhibiting viral replication in the human macrophages. DNA was extracted from HeLa cells infected with SINV and the DNA fragments separated through agarose gel electrophoreses. There were multiple bands visible in the infected samples whereas the negative control did not show multiple bands, only one large band of genomic DNA. The presence of multiple DNA fragments in infected cells and absence of those fragments from uninfected cells were suggestive of virus induced apoptosis.

South Africa is not excluded from the problems encountered world-wide in the treatment of nosocomial urinary tract infections, commonly caused by enzyme-producing Enterobacteriaceae. These enzymes include the Ã-lactamases and extended-spectrum Ã-lactamases (ESBLs) capable of hydrolysing the Ã-lactam agents and in particular the expanded-spectrum cephalosporins frequently used. The study was designed to determine the role of Ã-lactamases in resistance development in commonly encountered pathogens implicated in urinary tract infections and to characterise the enzymes involved. Resistance to the Ã-lactam agents amoxicillin, ceftriaxone, ceftriaxone, piperacillin and cefoxitin was suspected to involve the presence of one or more β-lactamases in the isolates from Bloemfontein hospitals. Diverse and complex β-lactamases were identified and ESBLs were detected in 80% of the isolates. These β-lactamases were characterised by isoelectric focusing (IEF) and genetic analysis (DNA amplification by PCR) to investigate the presence of possible genes responsible for resistance development. The production of blaTEM and blaSHV type genes was demonstrated. Isolates harbouring these genes were highly resistant to amoxicillin and piperacillin, with MIC90s of >128μg/ml. Resistance to these antibiotics was shown to be readily transferred between strains and there was an indication that the resistance genes are carried on plasmids and was transferred by conjugation. A plasmid of 9-10 kb was detected in 83% of the isolates and could be one of the mechanisms implicated in the transfer of ESBLs in uropathogenic bacteria. Ã-Lactam resistance could be attributed to the presence and action of Ã-lactamases such as the TEM and SHV type enzymes and this resistance can be transmitted between bacteria, causing problems specifically in the hospital environment. Further and continuous investigations are required to find a solution for this ever increasing problem.

Crimean-Congo haemorrhagic fever (CCHF) is a tick-borne viral zoonosis widely distributed in Africa, Asia, Russia and the Balkans. The causative agent, CCHF virus (CCHFV) has the propensity to cause nosocomial infections with a high fatality rate. Cases of CCHF are diagnosed annually in southern Africa. Increasing numbers of cases are seen in regions of Asia and in the past ten years CCHFV has emerged in several countries in the Balkans and re-emergence in south-western regions of the Russian Federation. Diagnosis of CCHFV infections during the acute phase is based on isolation of the virus or amplification of viral RNA. Patients that survive the infection have a demonstrable IgG and IgM antibody response, usually from day 5 to 7 after onset of illness. Current serological diagnostic assays based on ELISA or IF use inactivated virus which requires biosafety level 4 facilities for culturing the virus and therefore limits the number of laboratories that can prepare suitable reagents. Preparation of recombinant antigens would enable laboratories to perform serological diagnosis of CCHFV infections and surveillance studies. The purpose of this study was to prepare a recombinant CCHFV nucleoprotein using a bacterial expression system, to determine if the protein was immunogenic and to determine if the protein was able to detect IgG antibodies in survivors of CCHFV infection. The complete open reading frame of the gene encoding the NP of CCHFV was amplified by RT-PCR using primers specifically designed with restriction sites engineered to the primers to facilitate cloning. The amplicon was cloned into pGEM® T Easy vector using T/A cloning and the gene sequenced to confirm that the correct gene had been amplified and cloned into the vector for downstream cloning and expression applications. Initially we aimed to express the native gene using a bacterial expression system and the NP gene was rescued from the recombinant plasmid and cloned into pQE-80L vector using the BamH1 and Pst1 restriction sites present in the multiple cloning site on the vector. Various attempts were made to express the CCHFV NP protein however no protein was detectable using SDS PAGE methods or Western blot. The nucleotide sequence that we had determined for the open reading frame of our gene encoding the NP was analysed using the Rare Codon Analysis Tool software and we elected to codon optimize the gene for expression in E. coli. The optimized gene was synthesized by GenScript and supplied cloned in the multiple cloning site of pUC57. The optimized gene was excised from pUC57 and cloned into pColdTF bacterial expression vector. A 106 kDa protein was expressed from the construct likely representing the HIS tagged TF chaperone protein fused to the CCHFV NP protein and confirmed by Western blot analysis. A higher yield of the protein was present in the insoluble phase and as optimization of the growth and induction conditions did not significantly alter the insoluble to soluble ratio of the expressed protein, the protein was harvested from the insoluble phase by denaturing, purification and refolding of the protein. The biological activity of the recombinant protein was confirmed using immunoassays and by immunizing mice to determine if the antibodies induced by the recombinant protein could be detected using an antigen prepared from the whole virus. Four of five mice immunized with the recombinant NP had a detectable antibody response using an immunofluorescent assay. Serum samples from acute and convalescent patients collected at varying stages after onset of illness were reacted in a Western blot with the recombinant CCHFV NP protein. The recombinant antigen was able to detect IgG antibody in all the convalescent patient sera except two sera collected on days 14 and 15 during the acute phase. In contrast all the samples were detected using the recombinant antigen in an ELISA. Due to the potential biohazardous nature of samples only samples collected two weeks after onset of illness were tested. The results showed 100% concordance with the results obtained in an ELISA using mouse brain derived antigen. The assay was shown to be reproducible and stability studies showed that four months after preparation the protein was still active. A full validation of the protein using a large panel of serum samples from confirmed CCHF patients is now required. The results suggest that bacterially expressed proteins lacking post translational modifications and folding that occur with mammalian and baculovirus expression can be used in ELISA to detect IgG antibody against CCHFV in human sera which finds application in diagnostics, epidemiologic and surveillance studies.

To protect against sexual transmission, successful future HIV vaccines will likely be given to adolescents as a booster subsequent to primary immunization during infancy. In sub-Saharan Africa (SSA), a large proportion of children are chronically infected with a variety of helminths. These infections may suppress the ability of a host to elicit vaccine-induced Th1 responses that are considered important for a successful HIV vaccine. This study investigated the effect of chronic helminthic infection on the boosting capacity of a poxvirus-protein HIV vaccine regimen (SAAVI MVA-C and Env gp140 protein) in a mouse model. Groups of mice were prime-vaccinated with SAAVI MVA-C through an intramuscular injection, and Env gp140 protein formulated in Alum adjuvant which was administered via an intraperitoneal injection. These vaccinations were given concurrently, 2 weeks prior to infection with Schistosoma mansoni (Sm) through a percutaneous route. Control mice were either left uninfected (Naïve) or infected in the same manner (Sm) without vaccination. A booster vaccination was given 8 weeks post helminth infection. HIV-specific immune responses were analysed in the blood and spleens two weeks after booster vaccination. The magnitudes of cumulative IFN-γ ELISPOT responses to HIV Gag, RT and Env peptides were significantly (p<0.05) lower in the vaccinated and Sm-infected (Vaccine+Sm) mice (948 sfu/106) than vaccinated and uninfected (Vaccine) mice (1733 sfu/106), with IFN-γ responses to RT (CD8) being the most dominant for both mouse groups (Vaccine+Sm: 734 ± 221 sfu/106, Vaccine: 521 ± 116 sfu/106). No significant difference was observed in the magnitudes of cumulative IL-2 ELISPOT responses to the vaccine peptides between the Vaccine+Sm and Vaccine groups, however IL-2-producing T cell responses to Env (CD4) dominated in both mouse groups. Vaccine+Sm and Sm groups had similar IFN-γ- and IL-2-producing T cell responses to SEA. Splenocytes from Vaccine+Sm mice secreted less Th1 (IFN-γ, IL-2, TNF- α) and Th2 (IL-4, IL-6, IL-10) cytokines than those from uninfected vaccinated mice in response to HIV vaccine peptides. The total number of activated CD4+ T cells responding to vaccine peptides was greater for Vaccine+ Sm mice than Vaccine mice (p<0.05), however, no such statistical significance was observed in the differences seen between these vaccinated mouse groups for the number of activated CD8+ T cells. The frequencies of central memory activated CD4+ T cells were seen to be greater in Vaccine group (Gag; 34.28 ± 8.35%, Pol; 33.53 ± 6.34%, Env(CD4); 33.92 ± 3.87%, Env (CD8); 38.76 ± 10.52%) as opposed to the Vaccine+Sm group (Gag; 28.09 ± 3.95%, Pol; 26.45 ± 4.66%, Env (CD4); 28.79 ± 6.95%, Env (CD8); 28.65 ± 3.29%). Furthermore, Vaccine+Sm mice had higher titres of HIV-1 gp140- specific IgG1 antibodies (p<0.0001) (a Th2 antibody marker) but significantly less gp140- specific IgG2a (p<0.0001) and IgG2b (p<0.001) (Th1 antibody markers) antibodies. This trend was also observed with total non-Env-specific antibody titres. This study demonstrates that chronic helminthic infection is associated with an attenuated boosting capacity of a poxvirus-protein HIV vaccine in a mouse model, suppressing both T cell cytokine production and Th1-type antibody responses. Since HIV vaccine-induced Th1 responses are considered important for a successful HIV vaccine, these data suggest that chronic helminthiasis may impact negatively on future HIV vaccination outcomes in adolescents living in SSA where helminthic parasites are endemic.

Introduction: Nasopharyngeal (NP) colonization by Streptococcus pneumoniae (the pneumococcus) precedes the development of respiratory tract infection. Colonization by antimicrobial-resistant pneumococci, especially in infants, is a major public health concern as pneumococcus is a frequent cause of bacterial acute respiratory tract infections among children. This study longitudinally investigated antimicrobial resistance amongst pneumococci colonizing the nasopharynx of South African infants immunized with the 13- valent pneumococcal conjugate vaccine (PCV13). Furthermore, the study explored strainlevel pneumococcal colonization patterns and associated antimicrobial resistance determinants as well as the composition of the NP antibiotic resistome using shotgun metagenomic sequencing. Methods: NP swabs were collected every second week from birth through the first year of life from 137 infants who were immunized with 2+1 doses of PCV13. These were the first 137 infants enrolled in the cohort who had the most complete fortnightly NP sampling (defined as at least 23-26 fortnightly collected NP swabs). Pneumococci were identified and serotyped using conventional techniques, and their antibiotic susceptibility profiles determined by disc diffusion and E-test. A subset of 196 NP samples from 23 infants were selected based on changes in serotype or antimicrobial resistance. These were subjected to broth enrichment, total nucleic acid extraction and subsequent shotgun metagenomic sequencing. Sequence reads were assembled and aligned to reference pneumococcal genomes. In-silico pneumococcal capsular, multilocus sequence typing, and antimicrobial resistance determinants were described. Finally, antibiotic resistance genes were identified from all bacterial contigs, to determine the NP resistome. Results: 1520 pneumococcal (760 non-duplicate) isolates were recovered from 137 infants; including non-typeable (n = 99), PCV13 (n = 133), and non-PCV13 serotypes (n = 528). The prevalence of penicillin, erythromycin, and cotrimoxazole non-susceptibility was 19% (147/760; 95% CI 17-22%) (3% resistant), 18% (136/760; 95% CI 15-21%) (14% resistant) and 45% (344/760; 95% CI 42-49%) (36% resistant), respectively. The predominant penicillin-non-susceptible serotypes included 15B/15C (n = 20), 19A (n = 13), 15A (n = 10), 19F (n = 8), and 21 (n = 8). Multi-drug resistance (MDR) was observed in 9% (68/760; 95% CI 7-11%) of the isolates. PCV13 serotypes were more likely to be non-susceptible, compared to non-PCV13 serotypes, to penicillin (26% vs. 16%, p = 0.007), erythromycin (23% vs. 15%, p = 0.027) and cotrimoxazole (62% vs. 41%, p < 0.001). Non-susceptibility to penicillin, erythromycin, and cotrimoxazole remained relatively constant through the first year of life (X 2 test for trend: p = 0.184, range 0 – 25%; p = 0.171, range 0 – 27%; and p = 0.572, range 0 – 55%, respectively). Overall, penicillin or erythromycin-non-susceptible pneumococci were carried for a shorter duration than susceptible pneumococci (penicillin [mean days, 18 vs. 21, p = 0.013] and erythromycin [mean days, 18 vs. 21, p = 0.035]). Forty-five percentage (61/137) of infants carried the same serotype which acquired or lost resistance over time, and these changes were predominantly for penicillin (76%, 79/104). Of the 196 NP samples sequenced, 174 had corresponding positive cultures for pneumococci and, of these, 152 were assigned an in-silico serotype. Metagenomic sequencing detected a single pneumococcal serotype in 85% (129/152), and co-colonization in 15% (23/152) of NP samples, respectively. In total, 22 different pneumococcal serotypes were identified, with 15B/15C (n = 49) and 16F (n = 21) being the most common non-PCV13 serotypes, while 23F (n = 9) and 19A (n = 8) were the most common PCV13 serotypes. Twenty-six different sequence types (STs), including 4 novel STs were identified. Mutations in the folA and folP genes, associated with cotrimoxazole resistance, were detected in 89% (87/98) of cotrimoxazole-non-susceptible pneumococci and mutations in the pbp1a and pbp2x genes, known to confer beta-lactam resistance, were identified in penicillin nonsusceptible ST705215B/15C isolates. A total of 329 antimicrobial resistance (AMR) genes were detected in 64% (125/196) of the sequenced samples, including 36 non-redundant genes ranging from 1 to 14 genes per sample. The predominant AMR genes detected were those conferring resistance to beta-lactams (52%, 172/329), macrolide-lincosamide-streptogramin (17%, 56/329), and tetracycline (12%, 38/329). The msrD, ermB, and mefA genes were only detected from pneumococcal reads. The predominant resistance genes detected from nonpneumococcal reads included blaOXA-60, blaOXA-22, and blaBRO-1. Conclusion: NP carriage of antibiotic-non-susceptible pneumococci was relatively constant throughout the first year of life. Despite high vaccine coverage levels, PCV13 serotypes were identified and were more commonly non-susceptible to penicillin, erythromycin, and cotrimoxazole. Overall, penicillin or erythromycin-non-susceptible pneumococci were carried for a shorter duration than susceptible pneumococci, however, non-susceptible PCV13 serotypes were carried for a longer duration than non-susceptible non-PCV13 serotypes. Direct shotgun sequencing from enriched NP samples was shown to be a powerful technique for a detailed description of the pneumococcal component of the NP microbiome and resistome, and its use should be explored similarly for other bacteria in this niche.

Background: According to the World Health Organization Global TB report 2018, rifampicin monoresistant tuberculosis (RMR-TB) comprises 22% and 38% of all rifampicin-resistant TB (RR-TB) globally, and within South Africa, respectively. National surveys from South Africa show an increasing proportion of RMR-TB among TB cases compared to multi-drug resistant tuberculosis (MDR-TB) from 2001-02 (0.4% vs 2.9%) to 2012-14 (1.7% vs 2.8%). Data from the 2012-14 survey showed considerable variation in RMR-TB prevalence throughout the nine provinces of South Africa. Despite the above, factors associated with the rise in RMR-TB are unknown; and research is limited. This thesis aims to describe RMR-TB in more detail, by investigating the emergence and transmission of RR-TB strains in Khayelitsha, Western Cape Province, South Africa. This included: conducting a systematic review on temporal trends, transmission and risk factors associated with RMR-TB; describing the overall prevalence of RMR-TB among RR-TB; assessing the relative risk of RMR-TB versus MDR-TB among RRTB patients by HIV status during prior TB treatment; describing the distribution of rpoB mutations among RR-TB strains and assessing minimum inhibitory concentration (MIC) values in RR-TB strains with particular rifampicin-resistance (RIF R) conferring mutations; and investigating potential transmission through whole genome sequence (WGS) derived clusters among RR-TB strains. Methods: Routinely diagnosed RR-TB isolates are stored in a biobank at Stellenbosch University (SU). Clinical data (Médecins sans Frontières and additional data requested from the Western Cape Provincial Health Data Centre), together with stored RR-TB isolates from the biobank across 2013-15 inclusive were used to address research questions in Khayelitsha. To describe the overall prevalence of RMR-TB among all RR-TB over time, epidemiological data from 2008-17 were used. Laboratory techniques (sub-culturing of stored frozen cultures into mycobacterial growth indicator tubes [MGITs] for DNA extraction and quantitative phenotypic DST [q pDST]) involving the handling of live Mycobacterium tuberculosis cultures, were done in a Biosafety Level 3 laboratory at SU. Extracted DNA was sent to the University of Basel in Switzerland for library preparation and whole genome sequencing (WGS) on the Illumina HiSeq. The raw fastq WGS data files of the sequenced DNA were securely transferred to UCT. TB profiler was used to identify RIF R conferring mutations in rpoB and strain lineages. rpoB mutations were classified as high/moderate and minimal confidence in conferring rifampicin-resistance. q pDST (MGIT) was performed for isolates with minimal, moderate and rpoB confidence level mutations that were not classified. q pDST was also performed on isolates found to be rifampicin susceptible TB (RS-TB using WGS) [no RIF R conferring rpoB mutations detected] but were isolated from patients routinely diagnosed with RR-TB. A combination of software packages was used, as well as in-house developed scripts to compile a pipeline for WGS transmission cluster analysis. Computations were performed using facilities provided by UCTs ICTS High Performance Computing (HPC) team. Clusters were identified with Clusterpicker and by generating a single nucleotide polymorphism (SNP) distance matrix (SNP differences found between genomes) using R software; a SNP threshold of 12 was used to suggest recent transmission. Results: i) The overall prevalence of RMR-TB among all RR-TB remained relatively stable (17-31%) with no major temporal trend observed during 2008-17 in Khayelitsha. ii) The proportion of RMR-TB among all RR-TB was significantly higher among patients who were HIV positive during previous TB treatment compared to those who were HIV negative. iii) A high proportion (11%) of discordance was found among RIF R routinely diagnosed RR-TB patients (43% RMR-TB; 57% MDR-TB); resulting from possible mixed infections (43%), false-positive RIF R (18%) or both (39%). iv) The WGS-based DR-TB profile for rpoB mutations were distinctly different between RMR- and MDR-TB strains. The proportion of high/moderate vs minimal confidence levels for rpoB mutations was significantly higher among MDRTB (high confidence rpoB S531L mutation - Lineage 2) than for RMR-TB. v) Among RMR-TB strains, rpoB L511P (described as a disputed mutation, conferring minimal confidence for RIF R or low-level RIF R) was predominantly found among RMR-TB strains compared to MDR-TB strains. All rpoB L511P mutations (including RMR- and MDR-TB) tested phenotypically susceptible to rifampicin with MGIT; causing discrepancies between WGS and q pDST. All RMR-TB strains, including those with rpoB L511P mutations, had no other mutations conferring resistance to any of the other TB drugs. vi) Clustering was higher among MDR-TB strains compared to RMR-TB and more MDR-TB strains were clustered among strains with the rpoB S531L mutation compared to RMR-TB with the same mutation. In contrast, among strains with the rpoB L511P mutation, more RMR-TB strains were clustered compared to no clustering found among MDR-TB strains with the same mutation. Clinical data showed that RMRTB rpoB L511P clusters were due to closely related community acquired or nosocomial transmission; and SNP differences were < 5, suggesting direct transmission between RMR-TB rpoB L511P patients. Conclusion: In Khayelitsha, lower clustering of RMR-TB strains suggests reduced transmission compared to MDR-TB and along with a different rpoB mutation profile, suggests a different evolutionary mechanism of RIF R. Additionally, RMR-TB appears to be associated with HIV positivity during previous TB treatment, suggesting a role for HIV in the generation of RMR-TB. Given the association of rpoB L511P with low-level RIF R among RMR-TB strains, it is possible that different treatment approaches could be effective for these patients, as there are also no clinical trials to optimise treatment of these patients. WGS is beneficial, not only for understanding transmission of RR-TB strains, but also to be used in combination with MIC testing for individualised patient treatment regimens, in order to accurately diagnose RR-TB in future. Recommendations for preventing M.tb transmission; irrespective of HIV status; include early diagnosis and treatment initiation, and implementation of infection control programmes in various settings.

The estimated global burden of STEC (Shiga toxin producing Escherichia coli) is 2,481,511 illnesses, 269 deaths, and 26,827 DALYs with 48% of these being foodborne. This thesis provides information on STEC diagnostic strategy, undetected STEC in a tertiary referral hospital in Cape Town, and the virulence and antimicrobial resistance properties of tellurite resistant diarrheic E. coli isolated on CHROMagar(TM)STEC (CHROMagar Microbiology, Paris, France). Deploying the One - Health surveillance approach to study selected diarrheic bacterial pathogens in an informal settlement setting, this study sheds light on the extent of bacterial foodborne pathogens in human and non-human sources.