Defective Monocyte Enzymatic Function and an Inhibitory Immune Phenotype in Human Immunodeficiency Virus-Exposed Uninfected African Infants in the Era of Antiretroviral Therapy

Abstract Background Human immunodeficiency virus-exposed uninfected (HEU) infants are a rapidly expanding population in sub-Saharan Africa and are highly susceptible to encapsulated bacterial disease in the first year of life. The mechanism of this increased risk is still poorly understood. We investigated whether human immunodeficiency virus (HIV)-exposure dysregulates HEU immunity, vaccine-antibody production, and human herpes virus amplify this effect. Methods Thirty-four HIV-infected and 44 HIV-uninfected pregnant women were recruited into the birth cohort and observed up to 6 weeks of age; and then a subsequent 43 HIV-infected and 61 HIV-uninfected mother-infant pairs were recruited into a longitudinal infant cohort at either: 5–7 to 14–15; or 14–15 to 18–23 weeks of age. We compared monocyte function, innate and adaptive immune cell phenotype, and vaccine-induced antibody responses between HEU and HIV-unexposed uninfected (HU) infants. Results We demonstrate (1) altered monocyte phagosomal function and B-cell subset homeostasis and (2) lower vaccine-induced anti-Haemophilus influenzae type b (Hib) and anti-tetanus toxoid immunoglobulin G titers in HEU compared with HU infants. Human herpes virus infection was similar between HEU and HU infants. Conclusions In the era of antiretroviral therapy-mediated viral suppression, HIV exposure may dysregulate monocyte and B-cell function, during the vulnerable period of immune maturation. This may contribute to the high rates of invasive bacterial disease and pneumonia in HEU infants.

Human immunodeficiency virus-exposed uninfected (HEU) infants are particularly vulnerable to invasive bacterial disease [1,2], particularly, pneumonia [3] and diarrhea [4], and they have more frequent hospitalizations, more severe infections, and increased risk of treatment failure. However, the mechanism of this increased vulnerability remains unknown. The global population of HEU children is substantial, estimated at 1.2 million births annually, mainly within developing countries [5]. Therefore, a coordinated strategy is necessary to ensure their optimal health and wellbeing [6].
Due to the successful HIV test and treatment strategy globally, the prevalence of individuals receiving ART has risen considerably [5]. As a result, the number of HEU infants born to mothers receiving ART increased markedly; however, despite expanded implementation of prevention of vertical transmission programs, the risk of infection-related morbidity and mortality among HEU infants remains high [2], particularly the risk of encapsulated bacterial infection [22]. The immune profile among HEU infants is not well documented. Therefore, we addressed whether HIV-exposure and HHVs dysregulate infant immunity and/or the response to primary vaccination.

Study Design and Population
The study was conducted in Southern Malawi, at Ndirande Health Centre (primary healthcare facility in Blantyre) and at Queen Elizabeth Central Hospital (QECH) (tertiary teaching hospital in Blantyre). We recruited HEU and HU infant cohorts in 2 contiguous groups that were followed longitudinally preand postroutine childhood vaccination with pentavalent DPT-HepB-Hib immunization. The first group were aged 5-9 weeks (pre-1st vaccine dose), who were followed up to age 14-15 weeks (post-2nd vaccine dose), and the second group were aged 14-15 weeks (post-2nd vaccine dose), who were followed up to 18-23 weeks of age (post-3rd vaccine dose). We also recruited pregnant women in the early stages of labor at QECH maternity ward and subsequently their babies at birth (termed newborn birth cohort). Participating mothers were healthy (without disease), asymptomatic adults (≥18 years) comprising HIV-infected and HIV-uninfected volunteers. Human immunodeficiency virus testing was performed on maternal whole blood in the early stages of labor or at recruitment using 2 commercial point-of-care rapid HIV test kits, Determine HIV 1/2 kit (Abbott Diagnostic Division, Abbott Park, IL) and Unigold HIV 1/2 kit (Trinity Biotech Inc., Bray, Ireland), and had a CD4 + T-cell count performed. Human immunodeficiency virus-infected participants received first-line ART (Option B+ [Lamivudine, Tenofovir DF, and Efavirenz {3TC/TDF/EFV}]) during or prepregnancy. All HEU babies received nevirapine at birth for 6 weeks, followed by co-trimoxazole until 1 year of life. Exclusion criteria for the study participants were current or past history of smoking, heart disease, tuberculosis (TB), high blood pressure, drug use, syphilis, severe anemia (hemoglobin ,8 g/dL), placental abnormalities, infant prematurity, low birth weight or death, and existing comorbidities. All babies received Bacille Calmette-Guerin (BCG) vaccine at birth. Written informed consent was obtained from all participants before recruitment. Ethical approval was obtained from the University of Malawi College of Medicine Research and Ethics Committee (COMREC) (Blantyre, Malawi; protocol numbers P.11/11/1140, P.06/11/1088).

Sample Collection and Processing
We collected 5 mL venous blood from the infants, at 5-9, 14-15, and 18-23 weeks of age after attendance at the vaccination clinic for pentavalent DPT-HepB-Hib immunization. Participants with incomplete vaccine course were excluded. Mothers had 10 mL venous blood collected at the first visit and 3 mL breastmilk collected at all other time points. We collected up to 40 mL cord blood from the umbilical vein into sodium heparinized tubes using a 50-mL syringe, from the newborn interface of the placenta immediately after birth. Whole blood was kept at room temperature and processed within 2 hours. Peripheral blood mononuclear cells were isolated by density centrifugation. Plasma was separated by centrifugation at 1500 rpm for 10 minutes, aspirated, aliquoted, and stored at −80°C for later use. Breastmilk samples were collected by hand expression, fractionated into lipid and aqueous phase, and stored at −80°C. Due to limitations in the volume of blood collectable from very young babies and limited cell numbers, not all the assays were performed on every newborn or infant sample.

Human Immunodeficiency Virus Testing
Newborns' and infants' qualitative HIV deoxyribonucleic acid (DNA) polymerase chain reaction (PCR) tests were performed in batches of 23 at Malawi Liverpool Wellcome (MLW) core laboratories. Total DNA was isolated from 0.5 × 10 6 cells using AMPLICOR HIV-1 DNA test, V1.5 (Roche, Brea, USA)) according to the manufacturer's instructions. Human immunodeficiency virus DNA results in participant health passports were also collected.
Three HIV-RDT kits, from 2 separate manufacturers (Unigold and Determine), were used to confirm the presence or absence of HIV-specific immunoglobulins (Igs) in maternal peripheral blood. Cord blood mononuclear cells were stored in 500 µL RNAlater and analyzed at University College London (UCL) by digital PCR as described elsewhere [23].

Maternal CD4 Counts
Peripheral blood CD4 T-counts and full blood count were performed at the MLW Trust Clinical Research Programme Diagnostic Laboratory on an HmX analyzer (Beckman Coulter, Indiana, USA) using a standardized protocol. Blood was taken in the early stages of labor.

Phenotypic Analysis
Multicolor flow cytometry analysis was performed on whole blood. Samples were stained with the following fluorochrome-  Table 2). Samples were acquired on Beckman Coulter Cyan ADP and analyses were performed using FlowJo Version 7.6.5 and 10.5 software (TreeStar).

Measurement of Monocyte Phagosomal Enzymatic Activity
Phagosomal oxidative burst and bulk proteolytic function in monocytes was measured using a flow cytometry-based reporter bead assay as described previously [24] (Supplementary Methods 1A).

Human Cytomegalovirus Polymerase Chain Reaction
Real-time PCR was used to detect human CMV (hCMV) in HIV-infected and uninfected maternal breast milk and infant oropharyngeal throat swabs (Supplementary Methods 1C).

Detection of Cytomegalovirus-Specific Immunoglobulin (Ig)G and IgM Antibodies
Human CMV-specific IgM in HIV-infected and uninfected maternal and infant plasma was measured using a commercial enzyme-linked immunosorbent assay (ELISA) kit (IBL International, Hamburg, Germany) according to manufacturer's instructions. A semiquantitative, in-house hCMV IgG assay was used at the laboratories in the University of Birmingham, United Kingdom (Supplementary Methods 1D).

Epstein-Barr Virus Nested Polymerase Chain Reaction
A nested PCR measuring EBNA3B gene in HIV-infected and uninfected maternal breast milk and infant throat swabs were used (method as described in [26]).

Detection of Epstein-Barr Virus-Specific Immunoglobulin G Antibodies
Immunoglobulin G antibodies against EBV viral capsid antigen (VCA) were detected in plasma using a commercial ELISA kit (Diagnostic Automation, Woodland Hills, CA) according to the manufacturer's instructions.
Sandwich Enzyme-Linked Immunosorbent Assay to Detect Immunoglobulin G Specific to Vaccine Antigens In an in-house ELISA, tetanus toxoid (TT) or diphtheria toxoid (DT) (both National Institute for Biological Standards and Control [NIBSC] were tested, Potters bar, UK) (Supplementary Methods 1E). Optical density was measured (without acid stopping the reaction) after 10 minutes using an ELISA plate reader (Biotek, Cheshire, UK) set at 405 nm and SoftMax Pro software.

Statistical Analysis
Statistical analysis and graphical presentation were performed using Prism 7/8 (GraphPad Software, San Diego, CA), and Python (Python Software Foundation) was used to calculate summary statistics. Demographic and clinical characteristics were compared using Mann-Whitney U tests for continuous and Fisher's exact tests or χ 2 for discrete variables. The ELISpot data were reported as subtracted 2× background. Serotype-specific opsonophagocytic indexes (OPIs) were reported using geometric means and 95% confidence intervals. The OPIs were classified as being positive or negative based on the current recommended cutoff value of ,8 (negative) and ≥8 (positive). Results are reported as median and interquartile range (IQR) as stated.

Participant Characteristics
In the newborn birth cohort, 34 HIV-infected and 44 HIV-uninfected pregnant women were recruited: 2 newborns were excluded from the analysis due to death and HIV-positivity detected by digital droplet PCR. Human immunodeficiency virus-infected pregnant women received ART Option B+ (tenofovir/lamivudine/efavirenz) for an average of 18.7 (range, 1-143) months, with a mean nadir CD4 + T-cell count of 294 (range, 8-892) and were more likely to have had an elective caesarean birth, compared with HIV-uninfected pregnant women (P = .01). In the longitudinal infant cohort, 43 HIV-infected and 61 HIV-uninfected motherinfant pairs were recruited and sampled across time points, 5-9, 14-15, 18-23 weeks of age, corresponding to the Malawian routine infant vaccine schedule of the following: BCG at birth, then pentavalent vaccine at 6, 10, 14 weeks. Human immunodeficiency virus-infected women had received option B+ for an average of 9.28 (range, 1-72) months at the time of enrollment and had a mean nadir CD4 + T-cell count of 409 (range, 159-823).
There was no difference between maternal age or breastfeeding status, but mode of delivery was more often caesarean section in HIV-infected mothers compared with HIV-negative mothers ( Table 1).

Defective Monocyte Enzymatic Function
First, we used a flow cytometry-based whole blood phagocyte functional reporter bead assay [24] to assess monocyte function in cord blood from the birth cohort ( Figure 1A; Supplementary Figure 1). We assessed the ability of monocytes to internalize Alexa Fluor 405-labeled IgG-coated reporter beads at 1 hour post coincubation, as a proxy of uptake capacity. We showed that the proportion of monocytes that internalized reporter beads was similar between HEU infants and HU controls (56 [IQR,  vs 70 [IQR, 31-76]; P = .86) ( Figure 1B) (HU, n = 16; HEU, n = 12). We next assessed the phagosomal superoxide burst activity, and we found that it was lower in monocytes from HEU infants compared with HU controls (1 [IQR, 0.8-1.6] vs 3.9 [IQR, 2.7-8.9]; P = .0001) ( Figure 1C). Third, we assessed the phagosomal bulk proteolytic activity and showed that it was lower in monocytes from HEU newborns compared with HU controls (1.3 [IQR, 0.94-2.2] vs 2.8 [IQR, 2.1-3.3]; P = .0025) ( Figure 1D). Taken together, these data indicate altered monocyte phagosomal functional capacity at birth in HEU newborns.

Increased B-Cell Inhibitory Phenotype and Purified Protein Derivative Responses but Decreased T-Cell Phytohemagglutinin Responses
We next sought to investigate dysregulation in the adaptive arm of immunity using cord blood. We observed similar distributions of B-cell subsets between HEU and HU newborn babies (P . .1),  Next, using peripheral blood collected from the longitudinal infant cohort aged 5-9 weeks, we also found B-cell alterations in 2 subsets that are selectively dysregulated during chronic HIV infection [27,28]. The proportions of immature transitional and tissue-like memory B cells were lower in HEU infants than HU controls ( Figure 2A); however, the proportions of naive and central memory, CD4 + and CD8 + T-cell subsets were similar ( P . .1) (Supplementary Figure 2B and C).
When we looked at T cells from cord blood, the proportion of naive and central memory CD4 + and CD8 + T-cell subsets measured by CD45RA-CCR7 expression were similar between   Figure 2D). These data indicate that antigen-specific responses to PPD in HEU infants after BCG vaccination are enhanced.   Figure 6C). We next tested whether HIV exposure influences the levels of vaccine-induced functional antibody in infants aged 5-9 weeks. Using a Multiplex Opsonophagocytosis Assay, we measured opsonophagocytic activity of 13 vaccine serotypes from the PCV13 (1, 3, 4, 5, 6A, 6B, 7F, 9V, 14, 18C, 19A, 19F, 23F) in infant sera. We found no difference in the geometric mean OPI of 13 pneumococcal serotypes and geometric mean concentration (GMC) of serotype-specific IgG titers between HEU infants and HU controls ( Table 2).
To determine potential drivers of the described immune alterations in newborns and infants, we measured HIV Gag-specific responses in cord and peripheral blood mononuclear cells [29] using an 18-hour ex vivo IFNγ ELISpot assay in both the HEU newborn and the longitudinal infant cohorts ( Figure 5A

DISCUSSION
Human immunodeficiency virus-exposed but uninfected infants are at an increased risk of infectious disease even in the era of universal access to maternal ART; however, the underlying immunological basis is not well understood. We show altered monocyte phagosomal function, dysregulated B-cell homeostasis, and selective impairment of vaccine responses in HEU infants within the first 6 months of life. We also demonstrate evidence of HIV exposure and increased likelihood of hCMV exposure in HEU infants. We postulate that the variable severity and/or persistence of this immunological phenotype may explain the variable clinical manifestations reported in HEU infants [31,32], which may depend on the duration and intensity of exposure to HIV and other infectious cofactors.
The impaired monocyte phagosomal function in HEU newborns highlights their potential vulnerability to bacterial infection before the primary immunization series. Monocyte bactericidal activity requires uptake, reactive oxygen species formation, and phagosomes-lysosomes fusion resulting in inhibition, killing, and degradation of internalized bacteria [33]. In our setting, the validated flow cytometer reporter assay of phagocyte function that uses zymosan has shown poor immune function and superoxide burst activity in HIV-infected adults with active TB [34]. Monocyte functional impairment against encapsulated bacteria has also been observed in "age-associated" inflammation, where monocyte-activating cytokines tumor necrosis factor (TNF)-α and interleukin (IL)-6 are augmented [35]. Likewise, increased monocyte inflammatory markers, including sTNF-RI, IL-6, IP-10, oxLDL, and sCD14 are reported in HEU newborns [33], as well as enhanced proinflammatory cytokine secretion after stimulation with diverse pathogen-associated molecular pattern molecules at 6 weeks of age [36]. Moreover, recent PBMC transcriptomic profiling in HEU infants aged 1-2 years revealed down-regulated genes (LCN2, CAMP, HP, MMP8, BPI, LTF) associated with neutrophil function [37]. Taken together, this may explain HEU infant increased susceptibility to bacterial infection and pneumonia.
We also found increased FcLR4 expression on B cells. In chronically HIV-infected adults, the inhibitory BCR FcLR4 are overrepresented, an "exhausted" B-cell phenotype, with poor BCR-mediated activation and antigen-specific antibody production [38,39]. In HEU infants, at 6 to 14 weeks of age, we observed low proportions of tissue-like memory and immature-transitional B-cell subsets, which, conversely, are augmented in chronically HIV-infected adults [40,41] and are indicative of dysregulated B-cell homeostasis. More importantly, our data present evidence of increased exposure to HIV and hCMV in HEU infants from maternal HIV and/or viral proteins, hCMV recrudescence, and high PPD-specific IFNγ, which may promote B-cell dysregulation in early life. The PPD responses are shown to be increased in Mycobacterium tuberculosis-sensitized mothers, [42] and a bimodal response to BCG/PPD (high/low) has been reported in HEU infants in our setting [43]; BCG vaccination induces heterologous effects in myeloid cells at an epigenetic level in a process termed "trained immunity" [44], which may explain our findings. Taken together, the mechanisms of B-cell dysregulation are likely distinct from those seen in chronic HIV infection (hypergammaglobulinemia), due to the lack of replicative virus and preserved CD4 T cells. Altered B-cell homeostasis is associated with impaired antibody responses during chronic HIV infection [45,46]. Consistent with this observation, HIV-infected mothers in our cohort exhibited lower anti-Hib and anti-TT antibody titers than HIV-uninfected mothers using DPT-HepB-Hib. In agreement with maternal titers, we observed low anti-Hib titers in HEU infants. However, our observation is in contrast with . The IgG data were analyzed using Mann-Whitney U test reported as medians (interquartile range) and PCR data using a Fisher's exact test reporting effective size as a relative risk (RR). Blue circles are HU controls and red circles are HEU. Data are presented as means (standard deviation). DMSO, dimethyl sulfoxide. studies conducted in South Africa [47] and Uganda [48,49], who reported robust anti-Hib and anti-DT antibody titers in HEU infants. Differential vaccine immunogenicity is likely multifactorial, influenced by persistent immune exposure to HIV proteins, the time to maternal ART use, and the unique burden of infectious cofactors that likely contribute to a microenvironment of proinflammation. Consistent with previous observations [50], we also observed that a relatively large number of HEU infants mounted an IFNγ response after stimulation with HIV Gag; however, we did not detect IFNγ responses in HEU newborns, and responses were poor to the T-cell mitogen PHA (which cross-links the TCR/glycosylated surface proteins). Taken together, these data point towards HIV exposure as a possible driver of selective T-cell regulation at birth.
Our study limitations include the following: the limited number of assays per sample restricted adjustment for multiple comparisons; we excluded premature/low-birth weight/small for gestational age babies; no maternal viral load or clinical presentation of disease were taken; and HIV-negative mothers were not retested at the study end.

CONCLUSIONS
In conclusion, we show altered monocyte phagosomal function, dysregulated B-cell homeostasis, and selective impairment of vaccine antibody responses in HEU infants within the first 6 months of life. This period of vulnerability likely contributes to increased susceptibility to disease-causing bacteria that commonly cause life-threating illness such as pneumonia in HEU infants.

Supplementary Data
Supplementary materials are available at The Journal of Infectious Diseases online (http://jid.oxfordjournals.org/). Supplementary materials consist of data provided by the author that are published to benefit the reader. The posted materials are not copyedited. The contents of all supplementary data are the sole responsibility of the authors. Questions or messages regarding errors should be addressed to the author.