The Role of Viral Introductions in Sustaining Community-Based HIV Epidemics in Rural Uganda: Evidence from Spatial Clustering,Phylogenetics, and Egocentric Transmission Models

Background

It is often assumed that local sexual networks play a dominant role in HIV spread in sub-Saharan Africa. The aim of this study was to determine the extent to which continued HIV transmission in rural communities—home to two-thirds of the African population—is driven by intra-community sexual networks versus viral introductions from outside of communities.

Methods and Findings

We analyzed the spatial dynamics of HIV transmission in rural Rakai District, Uganda, using data from a cohort of 14,594 individuals within 46 communities. We applied spatial clustering statistics, viral phylogenetics, and probabilistic transmission models to quantify the relative contribution of viral introductions into communities versus community- and household-based transmission to HIV incidence. Individuals living in households with HIV-incident (n = 189) or HIV-prevalent (n = 1,597) persons were 3.2 (95% CI: 2.7–3.7) times more likely to be HIV infected themselves compared to the population in general, but spatial clustering outside of households was relatively weak and was confined to distances <500 m. Phylogenetic analyses of gag and env genes suggest that chains of transmission frequently cross community boundaries. A total of 95 phylogenetic clusters were identified, of which 44% (42/95) were two individuals sharing a household. Among the remaining clusters, 72% (38/53) crossed community boundaries. Using the locations of self-reported sexual partners, we estimate that 39% (95% CI: 34%–42%) of new viral transmissions occur within stable household partnerships, and that among those infected by extra-household sexual partners, 62% (95% CI: 55%–70%) are infected by sexual partners from outside their community. These results rely on the representativeness of the sample and the quality of self-reported partnership data and may not reflect HIV transmission patterns outside of Rakai.

Conclusions

Our findings suggest that HIV introductions into communities are common and account for a significant proportion of new HIV infections acquired outside of households in rural Uganda, though the extent to which this is true elsewhere in Africa remains unknown. Our results also suggest that HIV prevention efforts should be implemented at spatial scales broader than the community and should target key populations likely responsible for introductions into communities. Please see later in the article for the Editors'' Summary     

Effect of Synaptic Transmission on Viral Fitness in HIV Infection

HIV can spread through its target cell population either via cell-free transmission, or by cell-to-cell transmission, presumably through virological synapses. Synaptic transmission entails the transfer of tens to hundreds of viruses per synapse, a fraction of which successfully integrate into the target cell genome. It is currently not understood how synaptic transmission affects viral fitness. Using a mathematical model, we investigate how different synaptic transmission strategies, defined by the number of viruses passed per synapse, influence the basic reproductive ratio of the virus, R₀, and virus load. In the most basic scenario, the model suggests that R₀ is maximized if a single virus particle is transferred per synapse. R₀ decreases and the infection eventually cannot be maintained for larger numbers of transferred viruses, because multiple infection of the same cell wastes viruses that could otherwise enter uninfected cells. To explain the relatively large number of HIV copies transferred per synapse, we consider additional biological assumptions under which an intermediate number of viruses transferred per synapse could maximize R₀. These include an increased burst size in multiply infected cells, the saturation of anti-viral factors upon infection of cells, and rate limiting steps during the process of synapse formation.     

Incorporating metadata in HIV transmission network reconstruction: A machine learning feasibility assessment

HIV molecular epidemiology estimates the transmission patterns from clustering genetically similar viruses. The process involves connecting genetically similar genotyped viral sequences in the network implying epidemiological transmissions. This technique relies on genotype data which is collected only from HIV diagnosed and in-care populations and leaves many persons with HIV (PWH) who have no access to consistent care out of the tracking process. We use machine learning algorithms to learn the non-linear correlation patterns between patient metadata and transmissions between HIV-positive cases. This enables us to expand the transmission network reconstruction beyond the molecular network. We employed multiple commonly used supervised classification algorithms to analyze the San Diego Primary Infection Resource Consortium (PIRC) cohort dataset, consisting of genotypes and nearly 80 additional non-genetic features. First, we trained classification models to determine genetically unrelated individuals from related ones. Our results show that random forest and decision tree achieved over 80% in accuracy, precision, recall, and F1-score by only using a subset of meta-features including age, birth sex, sexual orientation, race, transmission category, estimated date of infection, and first viral load date besides genetic data. Additionally, both algorithms achieved approximately 80% sensitivity and specificity. The Area Under Curve (AUC) is reported 97% and 94% for random forest and decision tree classifiers respectively. Next, we extended the models to identify clusters of similar viral sequences. Support vector machine demonstrated one order of magnitude improvement in accuracy of assigning the sequences to the correct cluster compared to dummy uniform random classifier. These results confirm that metadata carries important information about the dynamics of HIV transmission as embedded in transmission clusters. Hence, novel computational approaches are needed to apply the non-trivial knowledge collected from inter-individual genetic information to metadata from PWH in order to expand the estimated transmissions. We note that feature extraction alone will not be effective in identifying patterns of transmission and will result in random clustering of the data, but its utilization in conjunction with genetic data and the right algorithm can contribute to the expansion of the reconstructed network beyond individuals with genetic data.     

Gaussian mixture clustering and imputation of microarray data

MOTIVATION: In microarray experiments, missing entries arise from blemishes on the chips. In large-scale studies, virtually every chip contains some missing entries and more than 90% of the genes are affected. Many analysis methods require a full set of data. Either those genes with missing entries are excluded, or the missing entries are filled with estimates prior to the analyses. This study compares methods of missing value estimation. RESULTS: Two evaluation metrics of imputation accuracy are employed. First, the root mean squared error measures the difference between the true values and the imputed values. Second, the number of mis-clustered genes measures the difference between clustering with true values and that with imputed values; it examines the bias introduced by imputation to clustering. The Gaussian mixture clustering with model averaging imputation is superior to all other imputation methods, according to both evaluation metrics, on both time-series (correlated) and non-time series (uncorrelated) data sets.     

Biological and molecular events associated with simultaneous transmission of plant viruses by invertebrate and fungal vectors

Viruses are likely to be the most dangerous parasites of living organisms because of their widespread occurrence, possible deleterious effects on their hosts and high rates of evolution. Virus host‐to‐host transmission is a critical step in the virus life cycle, because it enables survival in a given environment and efficient dissemination. As hosts of plant viruses are not mobile, these pathogens have adopted diverse transmission strategies involving various vector organisms, mainly arthropods, nematodes, fungi and protists. In nature, plants are often infected with more than one virus at a time, thereby creating potential sources for vectors to acquire and transmit simultaneously two or more viruses. Simultaneous transmission can result in multiple infections of new host plants, which become subsequent potential sources of the viruses, thus enhancing the spread of the diseases caused by these pathogens. Moreover, it can contribute to the maintenance of viral genetic diversity in the host communities. However, despite its possible significance, the problem of the simultaneous transmission of plant viruses by vectors has not been investigated in detail. In this review, the current knowledge on multiple viral transmissions by aphids, whiteflies, leafhoppers, planthoppers, nematodes and fungi is outlined.     

Human immunodeficiency virus type 1 env evolves toward ancestral states upon transmission to a new host

Selecting human immunodeficiency virus (HIV) sequences for inclusion within vaccines has been a difficult problem, as circulating HIV strains evolve relentlessly and become increasingly divergent over time. We report an assessment of this divergence from three perspectives: (i) across different hosts as a function of time of infection, (ii) between donors and recipients in known transmission pairs, and (iii) within individual hosts over time in relation to the initially replicating virus and to the deduced ancestral sequence of the intrahost viral population. Surprisingly, we consistently found less divergence between viruses from different individuals sampled in primary infection than in individuals sampled at more advanced stages of illness. Furthermore, longitudinal analysis of intrahost divergence revealed a 2- to 3-year period of evolution toward a common ancestral sequence at the start of infection, indicating that HIV recovers certain ancestral features when infecting a new host. These results have important implications for the study of HIV population genetics and rational vaccine design, including favoring the inclusion of viral gene sequences taken early in infection.     

Correcting bias due to missing stage data in the non-parametric estimation of stage-specific net survival for colorectal cancer using multiple imputation

BackgroundPopulation-based net survival by tumour stage at diagnosis is a key measure in cancer surveillance. Unfortunately, data on tumour stage are often missing for a non-negligible proportion of patients and the mechanism giving rise to the missingness is usually anything but completely at random. In this setting, restricting analysis to the subset of complete records gives typically biased results. Multiple imputation is a promising practical approach to the issues raised by the missing data, but its use in conjunction with the Pohar-Perme method for estimating net survival has not been formally evaluated.MethodsWe performed a resampling study using colorectal cancer population-based registry data to evaluate the ability of multiple imputation, used along with the Pohar-Perme method, to deliver unbiased estimates of stage-specific net survival and recover missing stage information. We created 1000 independent data sets, each containing 5000 patients. Stage data were then made missing at random under two scenarios (30% and 50% missingness).ResultsComplete records analysis showed substantial bias and poor confidence interval coverage. Across both scenarios our multiple imputation strategy virtually eliminated the bias and greatly improved confidence interval coverage.ConclusionsIn the presence of missing stage data complete records analysis often gives severely biased results. We showed that combining multiple imputation with the Pohar-Perme estimator provides a valid practical approach for the estimation of stage-specific colorectal cancer net survival. As usual, when the percentage of missing data is high the results should be interpreted cautiously and sensitivity analyses are recommended.     

Quantifying the impact of fixed effects modeling of clusters in multiple imputation for cluster randomized trials

In cluster randomized trials (CRTs), identifiable clusters rather than individuals are randomized to study groups. Resulting data often consist of a small number of clusters with correlated observations within a treatment group. Missing data often present a problem in the analysis of such trials, and multiple imputation (MI) has been used to create complete data sets, enabling subsequent analysis with well-established analysis methods for CRTs. We discuss strategies for accounting for clustering when multiply imputing a missing continuous outcome, focusing on estimation of the variance of group means as used in an adjusted t-test or ANOVA. These analysis procedures are congenial to (can be derived from) a mixed effects imputation model; however, this imputation procedure is not yet available in commercial statistical software. An alternative approach that is readily available and has been used in recent studies is to include fixed effects for cluster, but the impact of using this convenient method has not been studied. We show that under this imputation model the MI variance estimator is positively biased and that smaller intraclass correlations (ICCs) lead to larger overestimation of the MI variance. Analytical expressions for the bias of the variance estimator are derived in the case of data missing completely at random, and cases in which data are missing at random are illustrated through simulation. Finally, various imputation methods are applied to data from the Detroit Middle School Asthma Project, a recent school-based CRT, and differences in inference are compared.     

Cross-Species Virus Transmission and the Emergence of New Epidemic Diseases

Summary: Host range is a viral property reflecting natural hosts that are infected either as part of a principal transmission cycle or, less commonly, as “spillover” infections into alternative hosts. Rarely, viruses gain the ability to spread efficiently within a new host that was not previously exposed or susceptible. These transfers involve either increased exposure or the acquisition of variations that allow them to overcome barriers to infection of the new hosts. In these cases, devastating outbreaks can result. Steps involved in transfers of viruses to new hosts include contact between the virus and the host, infection of an initial individual leading to amplification and an outbreak, and the generation within the original or new host of viral variants that have the ability to spread efficiently between individuals in populations of the new host. Here we review what is known about host switching leading to viral emergence from known examples, considering the evolutionary mechanisms, virus-host interactions, host range barriers to infection, and processes that allow efficient host-to-host transmission in the new host population.     

Chemokine Receptor Utilization by Human Immunodeficiency Virus Type 1 Isolates That Replicate in Microglia

The role of human immunodeficiency virus (HIV) strain variability remains a key unanswered question in HIV dementia, a condition affecting around 20% of infected individuals. Several groups have shown that viruses within the central nervous system (CNS) of infected patients constitute an independently evolving subset of HIV strains. A potential explanation for the replication and sequestration of viruses within the CNS is the preferential use of certain chemokine receptors present in microglia. To determine the role of specific chemokine coreceptors in infection of adult microglial cells, we obtained a small panel of HIV type 1 brain isolates, as well as other HIV strains that replicate well in cultured microglial cells. These viruses and molecular clones of their envelopes were used in infections, in cell-to-cell fusion assays, and in the construction of pseudotypes. The results demonstrate the predominant use of CCR5, at least among the major coreceptors, with minor use of CCR3 and CXCR4 by some of the isolates or their envelope clones.     

Sensitivity analyses comparing outcomes only existing in a subset selected post-randomization, conditional on covariates, with application to HIV vaccine trials

In many experiments, researchers would like to compare between treatments and outcome that only exists in a subset of participants selected after randomization. For example, in preventive HIV vaccine efficacy trials it is of interest to determine whether randomization to vaccine causes lower HIV viral load, a quantity that only exists in participants who acquire HIV. To make a causal comparison and account for potential selection bias we propose a sensitivity analysis following the principal stratification framework set forth by Frangakis and Rubin (2002, Biometrics58, 21-29). Our goal is to assess the average causal effect of treatment assignment on viral load at a given baseline covariate level in the always infected principal stratum (those who would have been infected whether they had been assigned to vaccine or placebo). We assume stable unit treatment values (SUTVA), randomization, and that subjects randomized to the vaccine arm who became infected would also have become infected if randomized to the placebo arm (monotonicity). It is not known which of those subjects infected in the placebo arm are in the always infected principal stratum, but this can be modeled conditional on covariates, the observed viral load, and a specified sensitivity parameter. Under parametric regression models for viral load, we obtain maximum likelihood estimates of the average causal effect conditional on covariates and the sensitivity parameter. We apply our methods to the world's first phase III HIV vaccine trial.     

Interaction between HIV awareness, knowledge, safe sex practice and HIV prevalence: evidence from Botswana

This paper makes methodological and empirical contributions to the study of HIV in the context of Botswana, a country with high HIV prevalence. Comparable evidence is presented from India to put the Botswana results in perspective. The results point to the strong role played by affluence and education in increasing HIV knowledge, promoting safe sex and reducing HIV prevalence. The study presents African evidence on the role played by the empowerment of women in promoting safe sex practices such as condom use. The lack of significant association between HIV prevalence and safe sex practice points to the danger of HIV-infected individuals spreading the disease through multiple sex partners and unprotected sex. This danger is underlined by the finding that females with multiple sex partners are at higher risk of being infected with HIV. These results take on special policy significance in the context of Botswana, where the issue of multiple sex partners has not been adequately addressed in the programme to contain the spread of HIV.     

A signature in HIV-1 envelope leader peptide associated with transition from acute to chronic infection impacts envelope processing and infectivity

Mucosal transmission of the human immunodeficiency virus (HIV) results in a bottleneck in viral genetic diversity. Gnanakaran and colleagues used a computational strategy to identify signature amino acids at particular positions in Envelope that were associated either with transmitted sequences sampled very early in infection, or sequences sampled during chronic infection. Among the strongest signatures observed was an enrichment for the stable presence of histidine at position 12 at transmission and in early infection, and a recurrent loss of histidine at position 12 in chronic infection. This amino acid lies within the leader peptide of Envelope, a region of the protein that has been shown to influence envelope glycoprotein expression and virion infectivity. We show a strong association between a positively charged amino acid like histidine at position 12 in transmitted/founder viruses with more efficient trafficking of the nascent envelope polypeptide to the endoplasmic reticulum and higher steady-state glycoprotein expression compared to viruses that have a non-basic position 12 residue, a substitution that was enriched among viruses sampled from chronically infected individuals. When expressed in the context of other viral proteins, transmitted envelopes with a basic amino acid position 12 were incorporated at higher density into the virus and exhibited higher infectious titers than did non-signature envelopes. These results support the potential utility of using a computational approach to examine large viral sequence data sets for functional signatures and indicate the importance of Envelope expression levels for efficient HIV transmission.     

Mechanisms of enveloped RNA virus budding

To spread infection, enveloped viruses must bud from infected host cells. Recent research indicates that HIV and other enveloped RNA viruses bud by appropriating the cellular machinery that is normally used to create vesicles that bud into late endosomal compartments called multivesicular bodies. This new model of virus budding has many potential implications for cell biology and viral pathogenesis.     

Inflammatory Genital Infections Mitigate a Severe Genetic Bottleneck in Heterosexual Transmission of Subtype A and C HIV-1

The HIV-1 epidemic in sub-Saharan Africa is driven largely by heterosexual transmission of non-subtype B viruses, of which subtypes C and A are predominant. Previous studies of subtype B and subtype C transmission pairs have suggested that a single variant from the chronically infected partner can establish infection in their newly infected partner. However, in subtype A infected individuals from a sex worker cohort and subtype B individuals from STD clinics, infection was frequently established by multiple variants. This study examined over 1750 single-genome amplified viral sequences derived from epidemiologically linked subtype C and subtype A transmission pairs very early after infection. In 90% (18/20) of the pairs, HIV-1 infection is initiated by a single viral variant that is derived from the quasispecies of the transmitting partner. In addition, the virus initiating infection in individuals who were infected by someone other than their spouse was characterized to determine if genital infections mitigated the severe genetic bottleneck observed in a majority of epidemiologically linked heterosexual HIV-1 transmission events. In nearly 50% (3/7) of individuals infected by someone other than their spouse, multiple genetic variants from a single individual established infection. A statistically significant association was observed between infection by multiple genetic variants and an inflammatory genital infection in the newly infected individual. Thus, in the vast majority of HIV-1 transmission events in cohabiting heterosexual couples, a single genetic variant establishes infection. Nevertheless, this severe genetic bottleneck can be mitigated by the presence of inflammatory genital infections in the at risk partner, suggesting that this restriction on genetic diversity is imposed in large part by the mucosal barrier.     

Human immunodeficiency virus type 1 env sequences from Calcutta in eastern India: identification of features that distinguish subtype C sequences in India from other subtype C sequences

India is experiencing a rapid spread of human immunodeficiency virus type 1 (HIV-1), primarily through heterosexual transmission of subtype C viruses. To delineate the molecular features of HIV-1 circulating in India, we sequenced the V3-V4 region of viral env from 21 individuals attending an HIV clinic in Calcutta, the most populous city in the eastern part of the country, and analyzed these and the other Indian sequences in the HIV database. Twenty individuals were infected with viruses having a subtype C env, and one had viruses with a subtype A env. Analyses of 192 subtype C sequences that included one sequence for each subject from this study and from the HIV database revealed that almost all sequences from India, along with a small number from other countries, form a phylogenetically distinct lineage within subtype C, which we designate C(IN). Overall, C(IN) lineage sequences were more closely related to each other (level of diversity, 10.2%) than to subtype C sequences from Botswana, Burundi, South Africa, Tanzania, and Zimbabwe (range, 15.3 to 20.7%). Of the three positions identified as signature amino acid substitution sites for C(IN) sequences (K340E, K350A, and G429E), 56% of the C(IN) sequences contained all three amino acids while 87% of the sequences contained at least two of these substitutions. Among the non-C(IN) sequences, all three amino acids were present in 2%, while 22% contained two or more of these amino acids. These results suggest that much of the current Indian epidemic is descended from a single introduction into the country. Identification of conserved signature amino acid positions could assist epidemiologic tracking and has implications for the development of a vaccine against subtype C HIV-1 in India.     

HIV-1 Superinfection in the Antiretroviral Therapy Era: Are Seroconcordant Sexual Partners at Risk?

Background

Acquisition of more than one strain of human immunodeficiency virus type 1 (HIV-1) has been reported to occur both during and after primary infection, but the risks and repercussions of dual and superinfection are incompletely understood. In this study, we evaluated a longitudinal cohort of chronically HIV-infected men who were sexual partners to determine if individuals acquired their partners'' viral strains.

Methodology

Our cohort of HIV-positive men consisted of 8 couples that identified themselves as long-term sexual partners. Viral sequences were isolated from each subject and analyzed using phylogenetic methods. In addition, strain-specific PCR allowed us to search for partners'' viruses present at low levels. Finally, we used computational algorithms to evaluate for recombination between partners'' viral strains.

Principal Findings/Conclusions

All couples had at least one factor associated with increased risk for acquisition of new HIV strains during the study, including detectable plasma viral load, sexually transmitted infections, and unprotected sex. One subject was dually HIV-1 infected, but neither strain corresponded to that of his partner. Three couples'' sequences formed monophyletic clusters at the entry visit, with phylogenetic analysis suggesting that one member of the couple had acquired an HIV strain from his identified partner or that both had acquired it from the same source outside their partnership. The 5 remaining couples initially displayed no evidence of dual infection, using phylogenetic analysis and strain-specific PCR. However, in 1 of these couples, further analysis revealed recombinant viral strains with segments of viral genomes in one subject that may have derived from the enrolled partner. Thus, chronically HIV-1 infected individuals may become superinfected with additional HIV strains from their seroconcordant sexual partners. In some cases, HIV-1 superinfection may become apparent when recombinant viral strains are detected.     

An HIV Epidemic Model Based on Viral Load Dynamics: Value in Assessing Empirical Trends in HIV Virulence and Community Viral Load

Trends in HIV virulence have been monitored since the start of the AIDS pandemic, as studying HIV virulence informs our understanding of HIV epidemiology and pathogenesis. Here, we model changes in HIV virulence as a strictly evolutionary process, using set point viral load (SPVL) as a proxy, to make inferences about empirical SPVL trends from longitudinal HIV cohorts. We develop an agent-based epidemic model based on HIV viral load dynamics. The model contains functions for viral load and transmission, SPVL and disease progression, viral load trajectories in multiple stages of infection, and the heritability of SPVL across transmissions. We find that HIV virulence evolves to an intermediate level that balances infectiousness with longer infected lifespans, resulting in an optimal SPVL∼4.75 log₁₀ viral RNA copies/mL. Adaptive viral evolution may explain observed HIV virulence trends: our model produces SPVL trends with magnitudes that are broadly similar to empirical trends. With regard to variation among studies in empirical SPVL trends, results from our model suggest that variation may be explained by the specific epidemic context, e.g. the mean SPVL of the founding lineage or the age of the epidemic; or improvements in HIV screening and diagnosis that results in sampling biases. We also use our model to examine trends in community viral load, a population-level measure of HIV viral load that is thought to reflect a population''s overall transmission potential. We find that community viral load evolves in association with SPVL, in the absence of prevention programs such as antiretroviral therapy, and that the mean community viral load is not necessarily a strong predictor of HIV incidence.     

The differential infectivity and staged progression models for the transmission of HIV

Recent studies of HIV RNA in infected individuals show that viral levels vary widely between individuals and within the same individual over time. Individuals with higher viral loads during the chronic phase tend to develop AIDS more rapidly. If RNA levels are correlated with infectiousness, these variations explain puzzling results from HIV transmission studies and suggest that a small subset of infected people may be responsible for a disproportionate number of infections. We use two simple models to study the impact of variations in infectiousness. In the first model, we account for different levels of virus between individuals during the chronic phase of infection, and the increase in the average time from infection to AIDS that goes along with a decreased viral load. The second model follows the more standard hypothesis that infected individuals progress through a series of infection stages, with the infectiousness of a person depending upon his current disease stage. We derive and compare threshold conditions for the two models and find explicit formulas of their endemic equilibria. We show that formulas for both models can be put into a standard form, which allows for a clear interpretation. We define the relative impact of each group as the fraction of infections being caused by that group. We use these formulas and numerical simulations to examine the relative importance of different stages of infection and different chronic levels of virus to the spreading of the disease. The acute stage and the most infectious group both appear to have a disproportionate effect, especially on the early epidemic. Contact tracing to identify super-spreaders and alertness to the symptoms of acute HIV infection may both be needed to contain this epidemic.