Search

The increase in the number of human immunodeficiency virus-1 (HIV-1)-infected children is a direct consequence of the heterosexual spread of the disease to women and the growing number of HIV-positive i.v. drug users. It is not known how the majority of infants born to HIV-1-infected women escape HIV-1 infection, and, for those infected, the timing of HIV-1 transmission has yet to be determined. In addition, the role of maternal antibodies in the prevention of HIV-1 transmission to the fetus is unclear. We have previously demonstrated a correlation between vertical transmission and the absence of high-affinity/avidity antibodies to a peptide, KRI-HIGPGRAFYT, which corresponds to a region of the primary neutralizing domain of the gp120 V3 loop of HIVMN (MN-PND). The present study examines the correlation between the presence of these high affinity antibodies in women completing a pregnancy or undergoing an elective abortion and the detection of HIV-1 infection in their aborted fetuses. In several instances, transmission occurred despite high-affinity antibodies to the MN-PND. We have, therefore, evaluated the reactivity of sera to different MN-PND variants. In one infant born to a mother with high-affinity/avidity antibodies to KRI-HIGPGRAFYT (classic MN-PND), the infected baby developed antibodies to an MN-PND variant peptide against which his mother did not mount a humoral immune response during pregnancy. This finding indicates that fetal infection with MN-PND escape mutants arising during pregnancy may occur during a period when the mother is serologically negative.

Pediatric AIDS is increasing in frequency due to a rise in the number of human immunodeficiency virus type 1 (HIV-1)-infected women of childbearing age. Because outcome studies reveal that most children infected peripartum manifest HIV-1-related disease in the first year of life, intrauterine infection has been suspected. Fetal tissues from 23 second-trimester abortuses were examined. The presence of HIV-1 nucleic acid sequences was determined by the polymerase chain reaction and used to define infection of the fetus. By analysis of available tissues, 7 of 23 fetuses were infected, while control fetal tissue was negative. In situ hybridization for HIV-1 DNA showed that only 1 of 8 infected abortuses was positive, while all samples of noninfected tissues revealed no HIV-1 DNA. These studies indicate that maternofetal transmission of HIV-1 may occur in 30% of pregnancies (7/23) by the end of the second trimester.

Some neonates with congenital human immunodeficiency virus type 1 (HIV-1) infection exhibit immune dysregulation. This suggests that fetal CD4+ cells, possibly thymocytes, may be infected during gestation. If thymocytes are infected, this may result in a disruption of T-cell differentiation. To examine this hypothesis, normal human fetal thymocytes were established in tissue culture, characterized, and then exposed to HIV-1. On initial isolation, fetal thymocytes were analyzed for phenotypic markers by flow cytometry and assessed for T-cell function by mitogen-stimulated thymidine incorporation. The thymocytes comprised greater than 70% double positive (CD4+, CD8+) cells and responded to T- but not to B-cell mitogens. Thereafter, thymocytes were incubated in either tissue culture medium containing infectious HIV-1 or in control (HIV-free) medium. Infection of fetal thymocytes was determined by light and electron microscopy in combination with immunocytochemistry, molecular hybridization, and an infectious cell center (ICC) assay. After 1 week in culture, the thymocytes exposed to HIV-1 were positive by immunocytochemistry for the HIV-1-associated protein gp41. In addition, the presence of HIV-1 DNA was detected by molecular hybridization confirming infection of these cells. The ICC assay demonstrated the production of infectious HIV-1 particles and budding of mature virions was observed by electron microscopy. These studies demonstrate that human fetal thymocytes can be infected with HIV-1 in vitro and that this infection results in production of infectious virions. These results support the hypothesis that vertical transmission of HIV-1 in vivo may result in the infection of fetal thymocytes, which may contribute to postnatal HIV-1-associated pathologic conditions.

Passive immunity is conferred to the fetus by maternal antibodies, the majority of which are transported across the placenta during the third trimester of pregnancy. To determine the placental transport of anti-HIV-1 antibodies, serum from 5 women infected with human immunodeficiency virus (HIV) and their abortuses were examined for anti-HIV-1 antibodies. The gestational age of the abortuses ranged from 18 to 24 weeks and following polymerase chain reaction amplification, HIV-1 gag DNA was detected in tissue from 2 of the abortuses. The concentration of total IgG antibodies present in cord blood ranged from 2.9% to 12.5% of maternal levels. Antibodies directed against the envelope proteins, gp160 and gp120, the reverse transcriptase protein, p66, and the capsular protein, p24, were present in fetal and maternal serum. Although IgG1 was the predominant subclass antibody generated in response to HIV-1 proteins, IgG2, IgG3, and IgG4 directed against HIV-1 proteins were also detected. There were large differences in the antigens recognized by the antibodies produced in the mothers, and the IgG subclasses of the antibodies produced. HIV-1 proteins recognized by antibodies present in cord blood were similar to those recognized by paired maternal serum and IgG1, IgG2, IgG3 recognizing HIV-1 proteins were detected in fetal serum. However, there was a dichotomy in placental transport of IgG subclass antibodies to HIV-1 proteins. The role of these antibodies in prevention of vertical transmission of HIV-1 has yet to be determined.

The central nervous system (CNS) of 221 adults and 31 infants or children with the acquired immunodeficiency syndrome (AIDS) was examined with immunocytochemistry for infectious agents and for human immunodeficiency virus-1 (HIV-1) antigen (gp41). Since the major risk factor in this population was intravenous drug abuse, there were more female and pediatric patients than in other neuropathology autopsy series. Although children had a different spectrum of pathologic changes, including less frequent opportunistic infections, women did not differ from men in terms of types or incidence of opportunistic infections, vascular disease, neoplasia, and subacute AIDS encephalitis (SAE). Subacute AIDS encephalitis was detected in 26% of adult and 48% of pediatric brains. Immunocytochemical analysis of 100 adult and 20 pediatric brains revealed gp41 immunoreactivity in 78% and 40%, respectively. Virtually all adult brains with SAE had gp41 immunoreactivity in macrophages and microglia. Even brains with no significant pathology had frequent gp41 immunoreactivity, especially in the basal ganglia. In pediatric brains, including cases with SAE, gp41 immunoreactivity was less abundant, suggesting the possibility of latent infection or viral clearance. Spinal cords with vacuolar myelopathy or corticospinal tract degeneration had only rare gp41-positive cells. Brains from 16 aborted fetuses from HIV-1-seropositive women were all negative for gp41 immunoreactivity, but 12 brains were positive for HIV-1 by the polymerase chain reaction. These results may indicate that HIV-1 infection in fetal brains is below the limits of detection of immunocytochemistry. The differences noted between adults and children suggest that adults more often have productive CNS HIV-1 infection.

Neurological disease is a common finding in children with AIDS and in others without signs of disease but with evidence of congenital HIV-1 infection. To investigate the possibility that HIV-1 can infect fetal central nervous system (CNS) tissue and therefore possibly serve as the substrate for the abnormal neurodevelopment characteristic of pediatric AIDS, eight abortus CNS samples (one set of twins) from seven HIV-1-seropositive intravenous drug users (IVDUs) and eight control abortus CNS samples from eight HIV-1-seronegative IVDUs were analyzed for HIV-1 infection. HIV-1 nucleic acid was detected only after the use of polymerase chain reaction (PCR) in three of eight CNS samples from HIV-seropositive IVDUs but not in samples from seronegative subjects. In situ hybridization confirmed that HIV-1 DNA sequences were in cells in the CNS parenchyma of two of the three positive samples. This study demonstrates that HIV-1 can infect human fetal CNS tissue in vivo, but that the use of PCR may be necessary for its detection.

In order to determine if cardiac tissue from AIDS patients or patients with seropositivity to HIV-1 might be infected by HIV-1, portions of myocardium obtained postmortem were evaluated for HIV-1 DNA sequences. Cellular DNA was extracted and digested with EcoR1 and Southern blots were performed. One of three AIDS hearts was positive for HIV-1 DNA sequences without amplification, whereas two additional hearts were positive for HIV-1 DNA after amplification. Accordingly, other tissue from the heart positive for HIV-1 without amplification was studied by electron microscopy to localize HIV virions. Unexpectedly, large numbers of proliferating multilamellated membrane bodies were identified in myocytes, predominantly associated with mitochondria. Identical membrane bodies were found in two additional AIDS hearts, and in one heart from a patient with seropositivity to the AIDS virus, but in none of three similarly fixed controls. Immunocytochemistry for HIV core (p24) and envelope (gp120) antigens did not localize gold-labeled antibodies to the membrane bodies. We believe this membranopathy may be an HIV-1- or AIDS-specific abnormality of unknown etiology that may be related to the ultimate development of cardiomyopathy. In addition, our studies provide further support that HIV-1 may be present in AIDS hearts, although as yet we cannot state with certainty where the HIV-1 is located in these tissues.