HIV Lipodystrophy Case Definition using Artificial Neural Network Modelling

Objective

A case definition of HIV lipodystrophy has recently been developed from a combination of clinical, metabolic and imaging/body composition variables using logistic regression methods. We aimed to evaluate whether artificial neural networks could improve the diagnostic accuracy.

Methods

The database of the case-control Lipodystrophy Case Definition Study was split into 504 subjects (265 with and 239 without lipodystrophy) used for training and 284 independent subjects (152 with and 132 without lipodystrophy) used for validation. Back-propagation neural networks with one or two middle layers were trained and validated. Results were compared against logistic regression models using the same information.

Results

Neural networks using clinical variables only (41 items) achieved consistently superior performance than logistic regression in terms of specificity, overall accuracy and area under the ROC curve. Their average sensitivity and specificity were 72.4 and 71.2%, as compared with 73.0 and 62.9% for logistic regression, respectively (area under the ROC curve, 0.784 vs 0.748). The discriminating performance of the neural networks was largely unaffected when built excluding 13 parameters that patients may not have readily available. The average sensitivity and specificity of the neural networks remained the same when metabolic variables were also considered (total 60 items) without a clear advantage against logistic regression (overall accuracy 71.8%). The performance of networks considering also body composition variables was similar to that of logistic regression (overall accuracy 78.5% for both).

Conclusions

Neural networks may offer a means to improve the discriminating performance for HIV lipodystrophy, when only clinical data are available and a rapid approximate diagnostic decision is needed. In this context, information on metabolic parameters is apparently not helpful in improving the diagnosis of HIV lipodystrophy, unless imaging and body composition studies are also obtained.

Expression patterns of genes encoding endomembrane proteins support a reduced function of the Golgi in wheat endosperm during the onset of storage protein deposition

Wheat storage proteins are deposited in the vacuole of maturing endosperm cells by a novel pathway that is the result of protein body formation by the endoplasmic reticulum followed by autophagy into the central vacuole, bypassing the Golgi apparatus. This model predicts a reduced role of the Golgi in storage protein accumulation, which has been supported by electron microscopy observations. To study this issue further, wheat cDNAs encoding three distinct proteins of the endomembrane system were cloned and characterized. The proteins encoded were homologues (i) of the ER translocon component Sec61 alpha, (ii) the vacuolar sorting receptor BP-80 which is located in the Golgi and clathrin-coated prevacuole vesicles (CCV), and (iii) the Golgi COPI coatomer component COP alpha. During endosperm development, the levels of all three mRNAs were highest in young stages, before the onset of storage protein synthesis, and declined with seed maturation. However, the relative mRNA levels of BP-80/Sec61 alpha and the COP alpha/Sec61 alpha were lower during the onset of storage protein synthesis than at earlier stages of endosperm development. These results support previous studies, suggesting a reduced function of the Golgi apparatus in wheat storage protein transport and deposition.

The role of CD4+ T cell help and CD40 ligand in the in vitro expansion of HIV-1-specific memory cytotoxic CD8+ T cell responses

CD4(+) T cells have been shown to play a critical role in the maintenance of an effective anti-viral CD8(+) CTL response in murine models. Recent studies have demonstrated that CD4(+) T cells provide help to CTLs through ligation of the CD40 receptor on dendritic cells. The role of CD4(+) T cell help in the expansion of virus-specific CD8(+) memory T cell responses was examined in normal volunteers recently vaccinated to influenza and in HIV-1 infected individuals. In recently vaccinated normal volunteers, CD4(+) T cell help was required for optimal in vitro expansion of influenza-specific CTL responses. Also, CD40 ligand trimer (CD40LT) enhanced CTL responses and was able to completely substitute for CD4(+) T cell help in PBMCs from normal volunteers. In HIV-1 infection, CD4(+) T cell help was required for optimal expansion of HIV-1-specific memory CTL in vitro in 9 of 10 patients. CD40LT could enhance CTL in the absence of CD4(+) T cell help in the majority of patients; however, the degree of enhancement of CTL responses was variable such that, in some patients, CD40LT could not completely substitute for CD4(+) T cell help. In those HIV-1-infected patients who demonstrated poor responses to CD40LT, a dysfunction in circulating CD8(+) memory T cells was demonstrated, which was reversed by the addition of cytokines including IL-2. Finally, it was demonstrated that IL-15 produced by CD40LT-stimulated dendritic cells may be an additional mechanism by which CD40LT induces the expansion of memory CTL in CD4(+) T cell-depleted conditions, where IL-2 is lacking.

CD40 ligand trimer and IL-12 enhance peripheral blood mononuclear cells and CD4+ T cell proliferation and production of IFN-gamma in response to p24 antigen in HIV-infected individuals: potential contribution of anergy to HIV-specific unresponsiveness

It has been suggested that CD4+ T cell proliferative responses to HIV p24 Ag may be important in the control of HIV infection. However, these responses are minimal or absent in many HIV-infected individuals. Furthermore, while in vitro and in vivo responses to non-HIV recall Ags improve upon administration of highly active antiretroviral therapy, there does not appear to be a commensurate enhancement of HIV-specific immune responses. It is possible that CD4+ p24-specific T cells are deleted early in the course of infection. However, it is also possible that a discrete unresponsiveness, or anergy, contributes to the lack of proliferation to p24. To evaluate the possible contribution of unresponsiveness to the lack of CD4+ T cell proliferation to p24 in HIV-infected individuals, we attempted to overcome unresponsiveness. CD40 ligand trimer (CD40LT) and IL-12 significantly increased PBMC and CD4+ T cell proliferative responses to p24 Ag in HIV-infected, but not uninfected, individuals. No increase in proliferative response to CMV Ag was observed. CD40LT exerted its effect through B7-CD28-dependent and IL-12- and IL-15-independent mechanisms. Finally, the increase in proliferation with CD40LT and IL-12 was associated with an augmented production of IFN-gamma in most, but not all, individuals. These data suggest the possible contribution of HIV-specific unresponsiveness to the lack of CD4+ T cell proliferation to p24 Ag in HIV-infected individuals and that clonal deletion alone does not explain this phenomenon. They also indicate the potential for CD40LT and IL-12 as immune-based therapies for HIV infection.

Maintenance of large numbers of virus-specific CD8+ T cells in HIV-infected progressors and long-term nonprogressors

The virus-specific CD8+ T cell responses of 21 HIV-infected patients were studied including a unique cohort of long-term nonprogressors with low levels of plasma viral RNA and strong proliferative responses to HIV Ags. HIV-specific CD8+ T cell responses were studied by a combination of standard cytotoxic T cell (CTL) assays, MHC tetramers, and TCR repertoire analysis. The frequencies of CD8+ T cells specific to the majority of HIV gene products were measured by flow cytometric detection of intracellular IFN-gamma in response to HIV-vaccinia recombinant-infected autologous B cells. Very high frequencies (0.8-18.0%) of circulating CD8+ T cells were found to be HIV specific. High frequencies of HIV-specific CD8+ T cells were not limited to long-term nonprogressors with restriction of plasma virus. No correlation was found between the frequency of HIV-specific CD8+ T cells and levels of plasma viremia. In each case, the vast majority of cells (up to 17.2%) responded to gag-pol. Repertoire analysis showed these large numbers of Ag-specific cells were scattered throughout the repertoire and in the majority of cases not contained within large monoclonal expansions. These data demonstrate that high numbers of HIV-specific CD8+ T cells exist even in patients with high-level viremia and progressive disease. Further, they suggest that other qualitative parameters of the CD8+ T cell response may differentiate some patients with very low levels of plasma virus and nonprogressive disease.

CC-chemokines enhance the replication of T-tropic strains of HIV-1 in CD4(+) T cells: role of signal transduction

This study demonstrates that several CC-chemokines, including those that inhibit entry and replication of macrophage-tropic strains of HIV, increase the replication of T cell (T)-tropic strains in CD4(+) T cells. Enhancement of T-tropic HIV replication is observed at early stages of replication, requires signaling through inhibitory guanine nucleotide-binding regulatory (Gi) proteins, and is associated with increased cell surface colocalization of CD4 and the T-tropic HIV coreceptor CXCR4. These findings may further our understanding of the factors that influence the replication and spread of T-tropic strains of HIV in vivo and suggest that the use of cell signaling CC-chemokines as therapeutic agents for the purpose of limiting HIV replication in vivo should be approached with caution.

The role of dendritic cells in the infection of CD4+ T cells with the human immunodeficiency virus: use of dendritic cells from individuals homozygous for the delta32CCR5 allele as a model

Despite exposure to multiple strains of both macrophage (M)-tropic and T cell (T)-tropic HIV, primary infection is largely restricted to relatively homogeneous M-tropic virus. Since dendritic cells (DCs) play a pivotal role in the early events of HIV infection, several studies have focused on the role of DCs in this restriction. It has been proposed that DCs are more efficiently infected with M-tropic versus T-tropic viruses; however, the infectability of DCs and the relevance of their infectability for inducing productive infection is controversial. It has also been suggested that variability in DC expression of coreceptors for M-tropic versus T-tropic virus could explain the restriction in the transmitting virus. Using HIV-pulsed DCs from individuals with a homozygous deletion in the CCR5 gene as a human "knockout" model, we demonstrate that infection of DCs per se is not necessary to promulgate infection in CD4+ T cells. The data also suggest that transmission of HIV to CD4+ T cells is not dependent on DC coreceptor expression.

Natural killer cells from human immunodeficiency virus (HIV)-infected individuals are an important source of CC-chemokines and suppress HIV-1 entry and replication in vitro

Macrophage inflammatory protein (MIP)-1alpha, MIP-1beta, and RANTES (regulated on activation, normal T cell expressed and secreted), which are the natural ligands of the CC-chemokine receptor CCR5, inhibit replication of MT-2- negative strains of HIV-1 by interfering with the ability of these strains to utilize CCR5 as a coreceptor for entry in CD4(+) cells. The present study investigates the capacity of natural killer (NK) cells isolated from HIV-infected individuals to produce CC-chemokines and to suppress HIV replication in autologous, endogenously infected cells as well as to block entry of MT-2-negative HIV into the CD4(+) T cell line PM-1. NK cells freshly isolated from HIV-infected individuals had a high number of mRNA copies for MIP-1alpha and RANTES. NK cells produced significant amounts of RANTES, MIP-1alpha, and MIP-1beta constitutively, in response to stimulation with IL-2 alone and when they were performing their characteristic lytic activity (K562 killing). After CD16 cross-linking and stimulation with IL-2 or IL-15 NK cells produced CC-chemokines to levels comparable to those produced by anti-CD3-stimulated CD8(+) T cells. Furthermore, CD16 cross-linked NK cells suppressed (49-97%) viral replication in cocultures of autologous CD8/NK-depleted PBMC to a degree similar to that of PHA or anti-CD3-stimulated CD8(+) T cells. In 50% of patients tested, NK-mediated HIV suppression could be abrogated by neutralizing antibodies to MIP-1alpha, MIP-1beta and RANTES; in contrast, CD8(+) T cell-mediated suppression was not significantly overcome upon neutralization of CC-chemokines. Supernatants derived from cultures of CD16 cross-linked NK cells stimulated with IL-2 or IL-15 dramatically inhibited entry of a MT-2-negative strain of HIV, BaL, in the CD4(+)CCR5(+) PM-1 T cell line. These data suggest that activated NK cells may be an important source of CC-chemokines in vivo and may suppress HIV replication by CC-chemokine-mediated mechanisms in addition to classic NK-mediated lytic mechanisms.

HIV replication in CD4+ T cells of HIV-infected individuals is regulated by a balance between the viral suppressive effects of endogenous beta-chemokines and the viral inductive effects of other endogenous cytokines

This study demonstrates that the beta-chemokines macrophage inflammatory proteins 1 alpha and 1 beta (MIP-1 alpha and MIP-1 beta) and, RANTES (regulated on activation, normally T-cell expressed and secreted) inhibit human immunodeficiency virus (HIV) replication in anti-CD3 or recall antigen-stimulated peripheral blood mononuclear cells (PBMCs) of asymptomatic HIV-infected subjects. Significant levels of beta-chemokines were produced by both CD4+ and CD8+ PBMC subsets from HIV-infected individuals. Neutralization of endogenous MIP-1 alpha, MIP-1 beta, and RANTES did not rescue HIV replication in cultures to which greater than 10% CD8+ T cells had been added, indicating that the HIV suppressor activity of CD8+ T cells cannot be explained entirely by the beta-chemokines. However, significant enhancement of viral replication was observed upon neutralization of endogenous beta-chemokines in CD8-depleted or CD4+ PBMCs from most donors, particularly in cultures with low inducible levels of HIV production. In contrast, certain endogenous proinflammatory cytokines induced HIV replication in these same cells. These data suggest that the levels of HIV replication in CD4+ PBMC reflect the balance of the opposing effects of endogenous suppressive factors, such as the beta-chemokines, and HIV-inducing cytokines, such as tumor necrosis factor alpha and interleukin 1 beta.