Various Viral Compartments in HIV-1-Infected Mothers Contribute to In Utero Transmission of HIV-1

Abstract

Perinatal HIV transmission occurs in utero or intrapartum. The mechanisms and timing of transmission are not clearly understood. To compare the genetic sequences of the V3 envelope region of infant's plasma HIV to that of the mother's plasma, peripheral blood mononuclear cells (PBMC) and vaginal secretions, and correlate with timing of transmission. All 3 infants had a positive HIV PCR in the first days of life, thus classified as in utero infections. In the first mother–infant pair, two different variants were present in the infant, one correlating with maternal PBMC virus and highly homologous to virus from vaginal secretions and the other identical to sequences in maternal plasma. In the second pair, the infant plasma virus was similar to that of maternal PBMC. In the third pair, the cord blood and infant plasma virus were highly similar to maternal vaginal virus. The presence of more than one HIV variant from the maternal blood and from the vaginal compartment in the cord blood of infants presumably infected in utero could point to more than one episode of transmission or, alternatively, to transmission of PBMC virus.

Introduction

T he risk of in utero and intrapartum mother-to-child transmission of HIV-1 is well recognized and dramatic progress has been achieved in decreasing this risk with antiretroviral prophylactic regimens and, when indicated, elective caesarean delivery.¹ Transmission in utero is believed to be mediated mainly through maternal–fetal microtransfusions or placental infection, whereas transmission intrapartum predominantly through exposure of the infant's skin and mucous membranes to HIV in maternal blood or secretions in the birth canal.² An additional mechanism might be swallowing by the fetus of HIV-contaminated amniotic fluid during gestation or during labor.

There is evidence of increased risk of perinatal HIV transmission when HIV DNA is detected in maternal genital secretions.³ There is also evidence that the genital tract may comprise a distinct tissue compartment of HIV-1.^4
–6 However, there is very limited information on how the infant virus compares with those found in the two maternal compartments, blood and genital tract. Such sequence comparisons would be useful in deciphering the mechanisms and pathways of HIV transmission to the infant. If distinct viral sequences were found in maternal blood and in vaginal secretions, a comparison with the neonatal viral sequences might shed light on the source of infant virus. The only report, to our knowledge, of a comparison of infant and maternal genital canal viral strains is in a research letter by Tovanabutra et al. in 2007,⁷ who demonstrated a closer correlation of the infant peripheral blood mononuclear cells (PBMC) virus with that of maternal PBMC than that of maternal plasma or genital secretions in 7 of 8 mother–infant pairs examined.

Our objective was to elucidate the pathway of transmission of HIV-1 from mother to infant by comparing the infant's viral sequences with those found in maternal vaginal secretions or blood. We sequenced the variable V3 region of the env gp 120 gene of HIV-1 clones from plasma and PBMC, when available of three perinatally-infected infants and those in their mothers' plasma, PBMC, and vaginal secretions. In one case of a caesarean delivery, amniotic fluid was also available for analysis. The results were correlated with time of mother-to-child transmission of HIV, as determined by the time of first positive HIV PCR in the infant.

Methods

Subjects

The study subjects were three mother–infant pairs enrolled in the CDC-funded Mother-Infant Rapid Intervention at Delivery (MIRIAD) study.^8,9 This study implemented voluntary rapid HIV testing in labor and delivery units for women with undocumented HIV status during the current pregnancy at 17 US hospitals from 2001 to 2005. HIV-1 test results were positive for 52 out of 7,753 consenting tested participants. There were five documented cases of mother-to-child transmission of HIV among those women. Three of these pairs had specimens available for testing, and they are the subjects of this study. Specimens of maternal and infant plasma, PBMC (purified by Ficoll gradient), supernatant of vaginal secretions, and umbilical cord fetal blood had been collected from the mother–infant pairs in the MIRIAD study. In the case of an elective caesarean delivery (one case among the 3 women), amniotic fluid was collected following hysterotomy and prior to rupturing the membranes, using a catheter to puncture the membranes and collect the fluid. Two Dacron swabs were used to collect vaginal secretions from all four quadrants. Material from the swabs was resuspended in phosphate-buffered saline and centrifuged to separate cellular components from supernatant. The HIV infection status of the infants was determined by HIV-1 DNA PCR testing of blood collected at several time points during the infant's first year of life. Infants were defined as HIV-infected if two separate blood specimens drawn on different days (not cord blood) were both positive for HIV DNA. In this study, all maternal samples used for sequencing were from the time of delivery, while age of the infant when the plasma was collected was 4, 3, and 6 weeks of age for infants 1, 2, and 3, respectively. None of the infants were breastfed.

Viral quantitation, PCR, and V3 sequence analysis

Cell-free virions contained in aliquots of plasma or vaginal and amniotic fluid supernatants were collected by high-speed centrifugation (60,000 g for 1 h at 4°C), and viral RNA was purified using Trizol reagent (Invitrogen, Carlsbad, CA). The level of viral RNA in the samples was quantified as described.¹⁰ Briefly, one tenth of the purified RNA was added to a real-time, reverse-transcriptase Taq Man PCR assay using primers and probe that bind to a conserved region of pol. HIV copy number was determined by plotting the threshold cycle value (C_T) for each unknown on a standard curve generated from HIV-RNA standards. Inhibition of the PCR assay was monitored by the inclusion of an internal control. Viral levels were reported as HIV-1 RNA copies per mL of plasma or fluid.

For amplification of the HIV envelope V3 sequences, the remaining RNA prepared from samples was reverse-transcribed to cDNA with the TaqMan^TM Reverse Transcription kit (Applied Biosystems, Foster City, CA), in a reaction containing PCR Buffer II (50 mM KCl, 10 mM Tris HCl, pH 8.3), 5.0 mM MgCl_2, 0.5 mM dNTP, 2.5 μM random hexamers, 25 units Multiscribe reverse transcriptase, and sample RNA. Reaction mixtures were incubated for 15 min at 42°C, followed by 5 min at 95°C, and 5 min at 4°C. Genomic DNA was purified from frozen samples of PBMC using the Wizard Purification columns (Promega, Madison, WI).

The V3 region of the envelope gene was amplified in two to four independent assays by nested PCR reaction, using conserved primers (5’-AGCACAGTACAATGTACACATGG-3’ and SW2R – 5’-AATTTCTGGGTCCCCTCCTG-3’in round one; 5’-TGTTAAATGGCAGTCTAGCAGAA-3’ and 5’-TTCTGGGTCCCCTCCTGA GG-3’in round 2) to produce a 334 bp product (nucleotides 779 to 1113 of HIV_HXB) as described.¹¹ The resulting V3 PCR products were pooled and cloned using the pCR 2.1-TOPO vector system (Invitrogen), and plasmids containing appropriately sized inserts were sequenced. HIV -V3 sequences were analyzed using MacVector, version 7.2.3 (Accelrys, San Diego, CA), and % diversity over the 95 amino acid region was determined by comparison of paired maternal and infant sequences.

Results

The timing of first detection of HIV DNA (and time of first testing) in the infant 1, 2, and 3 was day of life 4, 2, and 1, respectively, after birth. In all cases, HIV-1 was of subtype B. None of the mothers were on antiretroviral therapy prior to labor, but all three were given intrapartum antiretroviral drugs (zidovudine in all three cases and, in addition, single dose nevirapine for mothers 1 and 2) once the HIV antibody test results became available. All infants received a 6-week course of zidovudine (in addition, infants 1 and 2 received single dose nevirapine). Maternal CD4+ T cell counts in the three pairs were 141, 289, and 241 cells/mm³, respectively. Duration of rupture of the amniotic membranes in the three pairs was 8 h, right at delivery, and 6 h, respectively. The HIV viral loads of the biologic fluids tested from the three mother-infant pairs are shown in Table 1.

Table 1.

Viral Loads (VL) in Different Maternal and Infant Compartments and Respective Time Points in Samples from Three HIV-Infected Mother–Infant Pairs, MIRIAD Study, 2001–2005

MIP #	Maternal plasma at delivery	Vaginal fluid (sup)/ml	Amniotic fluid/ml	Infant plasma–infant age	Cord blood
1	10,100	1,660	No sample	2,240–4 weeks	12,000
2	2,630	2,580	50	500–3 weeks	No sample
3	8,140	50	No sample	142,000–6 weeks	465

MIP, Mother–infant pair.

Viral loads were determined by real time PCR¹⁰ and are expressed as copies per ml of fluid. The assay has a sensitivity of 50 copies/ml.

Mother–infant pair 1

Four HIV env clones spanning the V3 region were generated and sequenced from maternal plasma, six from maternal PBMC, and seven from vaginal fluid supernatant, all obtained at the time of delivery. In addition, five clones from umbilical cord blood and nine clones from the baby's plasma at 4 weeks of life were sequenced. These sequences represent the predominant genotypes found in each compartment and their alignments are shown in Fig. 1A. Two viral variants were identified in the cord blood of the infant as well as the infant plasma sample obtained at 4 weeks of age. The infant plasma and cord blood variants were identical, suggesting limited diversification. One cord blood variant and the predominant variant found in infant plasma were identical to a variant found in maternal plasma and only 4–6% divergent to several of the diverse variants found in maternal vaginal secretions, plasma, and PBMC. The second cord blood and infant variant was identical to sequences found in maternal PBMC samples collected on the day of delivery, and were highly similar to some vaginal variants (1–2% divergent). Since viral species found in the vaginal secretions were highly similar to maternal PBMC quasi-species, the vaginal virus might have originated from a virus that was possibly found in the mother's plasma earlier during gestation, or is only present at low levels in plasma. Interestingly, this infant viral variant (clone Inf1-Pl e3, Inf1-Cb g3) has lost a V3 N-glycosylation site (position 6 of the V3 loop) and an acidic charge (D→N at position 29 of the V3 loop), suggestive of a tropism switch from CCR5 to CXCR4.¹²

FIG. 1.

Alignments of the deduced envelope V3-region amino acid sequences amplified from available mother (M) and infant (Inf ) samples of plasma (Pl), vaginal fluid supernatants (Vs), cord blood plasma (Cb), amniotic fluid supernatant (Af ), and PBMC (C) in three mother–infant pairs, participants of the MIRIAD Study, 2001–2005. HIV sequences of V3 regions were amplified by RT-PCR using RNA purified from plasma, vaginal fluid, cord blood, and amniotic fluid, and PCR amplification of purified DNA from PBMC samples. Dots indicate identity to amino acid shown in the first line, and dashes indicate gaps introduced to maintain alignment. The number of clones containing sequences identical to variant shown are indicated in the column at the right. The asterisks identify the cysteines at the V3 loop boundaries. (A) contains sequences amplified from cord blood plasma and maternal samples obtained on the day of birth from Pair 1, as well as infant plasma obtained at 4 weeks of age. (B) contains sequences amplified from amniotic fluid, infant PBMC and maternal samples obtained on the day of birth from Pair 2, as well as infant plasma obtained at 3 weeks of age. (C) contains sequences amplified from cord blood plasma and maternal samples obtained on the day of birth from Pair 3, as well as infant plasma obtained at 6 weeks of age.

The infant was born vaginally and had an early positive HIV DNA test at 4 days of life (cord blood was also HIV DNA positive), thought to be consistent with in utero transmission.¹³

Mother–infant pair 2

Thirteen clones were generated from maternal plasma, five from maternal PBMC, and 10 from vaginal fluid supernatant at the time of delivery. Eight clones were sequenced from the infant's plasma at 3 weeks of age, and six from the infant's PBMC at the time of birth. In addition, nine clones were generated from amniotic fluid (Fig. 1B). Viral genotypes amplified from PBMC DNA of the infant on the day of birth, as well as plasma variants amplified from the infant at 3 weeks of age were all homologous and identical to viral variants found in amniotic fluid and maternal PBMC on the day of delivery. The predominant viral variants found in maternal plasma and vaginal secretions were more diverse (2–8% divergent from sequence in the infant), but the majority of maternal variants were highly similar to infant variants (5% or less divergent from the infant sequence). Given that the infant V-3 genotype at 3 weeks of life was identical to that of amniotic fluid, viral diversification in the infant was not observed. This infant was born by elective cesarean section and the first positive HIV PCR was at 3 days of life, thought as indicative of in utero infection,¹³ which is supported by the viral sequence comparisons.

Mother–infant pair 3

Five clones were generated from maternal plasma and 10 from vaginal fluid supernatant; eight clones were generated from infant plasma (tested at 6 weeks of life), and nine from umbilical cord blood plasma (Fig. 1C). PBMC (from mother and infant) were not available. All viral V3-variants from the infant's plasma were identical to sequences in maternal vaginal fluid collected on the day of birth and were less than 6% divergent from the majority of maternal plasma sequences collected at the time of delivery. The cord blood plasma sequences had a very high degree of homology to infant plasma genotypes and vaginal fluid sequences. This infant was born vaginally and the early positive HIV DNA test (at first day of life) would indicate that the infant was infected in utero.¹³

Discussion

In order to examine the mechanisms and origin of virus in mother-to-child transmission of HIV-1, we have compared the genetic sequences among viral populations detected at the time of delivery in three HIV-infected mothers and their infants with presumed in utero HIV infection. All these mothers had undocumented HIV status during their pregnancy and had not received antiretroviral prophylaxis prior to their labor and delivery. We compared the env V3 regions of virus from maternal PBMC, plasma, and vaginal fluid and their perinatally infected infants'cord blood and plasma (in one case also the amniotic fluid). Four main points emerge from our results. First, the degree of viral diversity in infant blood plasma and of viral diversification during the first few weeks of the infected infant's life is small. However, there can be more than one circulating HIV variant in the perinatally-infected newborn's plasma or PBMC. Second, the infant may be infected with more than one variant that can be identified in different maternal sources (circulating PMBC, plasma, or the vaginal compartment) and may have been acquired at different times. Third, the infant virus was more closely related to the maternal PBMC rather than plasma cell-free virus in two mother–infant pairs where this was evaluated. In fact, the infant plasma virus was not always detected in maternal plasma at the time of birth. Fourth, cord blood sampling might be a useful aid in deciphering the origin of HIV transmitted to the infant.

Sequence analyses of HIV-1 variants isolated from infected individuals have revealed a highly homogeneous population within the HIV-1 env gene early following transmission. ^14

–18 This homogeneity extends to other tissue compartments.¹⁶ In chronically infected individuals, on the other hand, heterogeneous and distinct viral populations emerge in different compartments.^15,19 The cervicovaginal compartment can harbor genetically diversified HIV-1 variants distinguishable from those present in blood. ^4,5 It has been argued that this reflects a gradual diversification of a common dominant viral variant.

With regards to perinatal HIV transmission, most evidence supports restricted viral heterogeneity in blood near the time of infection among the majority of infected infants.^10,20,21 However, transmission of multiple variants has also been described.^{22

–29} Moreover, transmission at the time of delivery has been primarily associated with a single viral quasi-species^11,29,30 that is a minor maternal variant,²⁹ whereas in utero transmitters were more likely to transmit single or multiple major maternal viral variants. ^23,28,30 All these findings are supported by our results. Of interest, one of the viruses in infant 1 has the hallmarks of CXCR4 receptor utilization, an infrequent finding in mother to child HIV transmission, and usually associated with more advanced disease and a more aggressive course.³²

Two of the mother–infant pairs tested in this study, pairs 1 and 3, both presumably with infants infected in utero, had a circulating viral strain in the cord blood and in infant's blood that corresponded with a strain found in vaginal secretions. In mother–infant pair 3 in particular, since this was the only strain detected in the infant, in utero infection could be possible if the infant was infected with a cell-associated maternal virus not persisting or present at low-abundance in the plasma at the time of delivery but still present in the vaginal compartment. Alternatively, intrapartum transmission would have to be assumed, which led, however, to a very rapid positive PCR test result. The very rapid detection of this virus in the baby's blood would be difficult to explain unless the lymphocytes are activated (possibly by earlier in utero transmission of another strain of HIV or from another stimulus) and thus provide a substrate for very rapid HIV proliferation. Our previous findings with L-selectin, an activation marker, in the plasma of HIV-infected newborns had suggested that lack of a positive HIV test in the first few days of life correlated with a lack of lymphocytic activation.³³ It has also been speculated that a small degree of diversity between HIV quasi-species in mother's blood and cord blood is related to late in utero transmission.³⁴ We and others have previously argued that most HIV intrauterine transmission occurs at the very late stages of pregnancy.^2,35

The results of mother–infant pair 1 are also compatible with transmission of virus on more than one occasion, during gestation (the virus from maternal blood) and at delivery (the virus from the vaginal compartment). Such a conclusion has been reached by other investigators who examined cell-associated viral variants in mothers and their infected infants.^22,23,31 Cell-associated virus in the blood and cervicovaginal secretions of the mother was more closely related to transmission of infection to the infant than plasma cell-free virus,^3,36 which is also highly supported by our results.

Our study has the limitation of a small number of mother–infant pairs. However, given the dramatic decrease in the rate of mother-to-child transmission of HIV in the US, obtaining a larger number of mothers with infected infants presents logistical difficulties, particularly in view of the detailed set of samples from different maternal compartments that is required for such an analysis. Another limitation of this study is the small number of clones studied; it is possible that they underestimate the viral diversity present. We examined only the V3 region of env; it could be argued that sampling larger regions might show more diversity; V3 is, however, one of the most heterogeneous regions of the HIV genome. The limited sample material available for analysis and the relatively low viral copy number restricted the amount of RNA and DNA available for PCR amplification, cloning, and more extensive sequencing. Notwithstanding these limitations, this study provides the unique angle of a comparison of the infant virus not only with virus in maternal blood, but also in the vaginal canal. Additionally, we were able to show identity between V3 genotypes in infant plasma and maternal samples obtained from different compartments. Our findings point to transmission of cell-associated maternal virus during late gestation and may suggest more than one episode and time of mother-to-child transmission of HIV. Shedding further light on the mechanisms of transmission of HIV from mother-to-infant is helpful in optimizing prevention efforts and making them more targeted and precise.

Footnotes

Acknowledgment

The MIRIAD study was coordinated and funded by the National Center for HIV, STD, and TB Prevention at CDC under cooperative agreements U64/217724, 417719, 417735, 517715, and 617734. We thank all the MIRIAD study investigators and the many people who provided critical input into protocol development, training at the sites, and guidance and support throughout the duration of the study. We especially thank Angela Bradley–Byers, RN (New Orleans, Louisiana) and Yvette Rivero, BA (Miami, Florida) for their superb coordination at the hospitals where the delivery of the 3 infants in our study occurred. Drs. Mary Glenn Fowler and Alan Greenberg (CDC) provided scientific guidance and support throughout the study. We thank Margaret Lampe, RN, Rosalind Carter, PhD, Yolanda Olszewski, MPH, Renata Dennis, RN, MPH, Steven Nesheim, MD, Mardge Cohen, MD, Mayris Webber, DrPH, Bernard Branson, MD, Michael Lindsay, MD, MPH, Andre Nahmias, MD, Francis Lee, PhD, Pat Garcia, MD, Elaine Abrams, MD, Gwendolyn Scott, MD, Clyde Hart, PhD, Jeffrey Wiener, MS and Sanjyot Shinde, PhD, for their tireless work on the MIRIAD study and Drs. Chin–Yih Ou and Jeff Johnson for their helpful comments after reading a draft of this manuscript.

Author Disclosure Statement

The authors do not have any conflicts of interest to declare.

The findings and views expressed in this paper are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention.

References

Jamieson

, Clark

, Kourtis

et al. Recommendations for human immunodeficiency virus screening, prophylaxis, and treatment for pregnant women in the United States. Am J Obstet Gynecol, 2007; 197:S26–32.

Kourtis

, Bulterys

, Nesheim

, Lee

. Understanding the timing of HIV transmission from mother to infant. JAMA, 2001; 285:709–712.

Tuomala

, O'Driscoll

, Bremer

et al. Cell-associated genital tract virus and vertical transmission of HIV type 1 in antiretroviral-experienced women. J Infect Dis, 2003; 187:375–384.

Ellenbrock

, Lennox

, Clancy

et al. Cellular replication of HIV type 1 occurs in vaginal secretions. J Infect Dis, 2001; 184:28–36.

Overbaugh

, Anderson

, Ndinya–Achola

, Kreiss

. Distinct but related HIV-1 variant populations in genital secretions and blood. AIDS Res Hum Retrov, 1996; 12:107–115.

Kliks

, Contag

, Corliss

et al. Genetic analysis of viral variants selected in transmission of HIV to newborns. AIDS Res Hum Retrovir, 2000; 16:1223–1233.

Tovanabutra

, de Souza

, Sittisombut

et al. HIV-1 genetic diversity and compartmentalization in mother/infant pairs infected I CRF01-AE. AIDS, 2007; 21:1050–1053.

Bulterys

, Jamieson

, O'Sullivan

et al. Rapid HIV-1 testing during labor. A multicenter study. JAMA, 2004; 292:219–223.

Jamieson

, Cohen

, Maupin

et al. Rapid human immunodeficiency virus-1 testing on labor and delivery in 17 US hospitals: The MIRIAD experience. Am J Ob Gyn, 2007; 197:S72–82.

10.

Henning

, Lacour

, Amedee

. Efficient methodologies for sensitive HIV-1 RNA quantitation from plasma and vaginal secretions. J Clin Virol, 2009; 46:309–313.

11.

Wolinsky

, Wike

, Korber

et al. Selective transmission of human immunodeficiency virus type-1 variants from mothers to infants. Science, 1992; 255:1134–1137.

12.

Pollakis

, Kang

, Kliphuis

, Chalaby

, Goudsmit

, Paxton

. N-Linked glycosylation of the HIV Type-1 gp120 envelope glycoprotein as a major determinant of CCR5 and CXCR4 coreceptor utilization. J Biol Chem, 2001; 276:13433–13441.

13.

Bryson

, Luzuriaga

, Sullivan

, Wara

. Proposed definitions for in utero versus intrapartum transmission of HIV-1. N Engl J Med, 1992; 327:1246–1247.

14.

Richardson

, Mbori–Ngacha

, Lavreys

et al. Comparison of HIV-1 viral loads in Kenyan women, men and infants during primary and early infection. J Virol, 2003; 77:7120–7123.

15.

Poss

, Rodrigo

, Gosink

et al. Evolution of envelope sequences from the genital tract and peripheral blood of women infected with clade A HIV-1. J Virol, 1998; 72:8240–8251.

16.

Freel

, Fiscus

, Pilcher

et al. Envelope diversity, coreceptor usage and syncytium-inducing phenotype of HIV-1 variants in saliva and blood during primary infection. AIDS, 2003; 17:2025–2033.

17.

Zhang

, MacKenzie

, Cleland

, Holmes

, Brown

, Simmonds

. Selection for specific sequences in the external envelope protein of HIV-1 upon primary infection. J Virol, 1993; 67:3345–3356.

18.

Zhu

, Mo

, Wang

et al. Genotypic and phenotypic characterization of HIV-1 in patients with primary infection. Science, 1993; 261:1179–1181.

19.

Ohagen

, Devitt

, Kunstman

et al. Genetic and functional analysis of full-length HIV-1 env genes derived from brain and blood of patients with AIDS. J Virol, 2003; 12336–45.

20.

Mulder–Kampinga

, Kuiken

, Dekker

et al. Genomic HIV-1 RNA variation in mother and child following intra-uterine virus transmission. J Gen Virol, 1993; 74:1747–1756.

21.

Zhang

, Orti

, Du

et al. Phylogenetic and phenotypic analysis of HIV type 1 Env gp120 in cases of subtype C mother-to-child transmission. AIDS Res Hum Retrovir, 2002; 18:1415–1423.

22.

Briant

, Wade

, Puel

, Leigh Brown

, Guyader

. Analysis of envelope sequence variants suggests multiple mechanisms of mother-to-child transmission of HIV-1. J Virol, 1995; 69:3778–3788.

23.

Nowak

, Karlsson

, Naver

et al. The selection and evolution of viral quasi-species in HIV-1 infected children. HIV Med, 2002; 3:1–11.

24.

Renjifo

, Chung

, Gilbert

et al. In-utero transmission of quasi-species among HIV type 1 genotypes. Virol, 2003; 307:278–282.

25.

Scarlatti

, Leitner

, Halapi

et al. Comparison of variable region 3 sequences of HIV-1 from infected children with the RNA and DNA sequences of the virus population of their mothers. Proc Natl Acad Sci USA, 1993; 90:1721–1725.

26.

Pasquier

, Cayrou

, Blancher

et al. Molecular evidence for mother-to-child transmission of multiple variants by analysis of RNA and DNA sequences of HIV-1. J Virol, 1998; 72:8493–8501.

27.

Verhofstede

, Demecheleer

, De Cabooter

et al. Diversity of the HIV-1 env sequence after vertical transmission in mother-child pairs infected with HIV-1 suptype A. J Virol, 2003; 77:3050–3057.

28.

Lamers

, Sleasman

, She

et al. Persistence of multiple maternal genotypes of HIV-1 in infants infected by vertical transmission. J Clin Invest, 1994; 93:380–390.

29.

Dickover

, Garratty

, Plaeger

, Bryson

. Perinatal transmission of major, minor, and multiple maternal human immunodeficiency virus type 1 variants in utero and intrapartum. J Virol, 2001; 75:2194–2203.

30.

Ahmad

, Baroudy

, Baker

, Chappey

. Genetic analysis of HIV-1 envelope V3 region isolates from mothers and infants after perinatal transmission. J Virol, 1995; 69:1001–1012.

31.

Kwiek

, Russell

, Dang

et al. The molecular epidemiology of HIV-1 envelope diversity during HIV-1 subtype C vertical transmission in Malawian mother-infant pairs. AIDS, 2008; 22:863–871.

32.

Derdeyn

, Hunter

. Viral characteristics of transmitted HIV. Curr Opin HIV AIDS, 2008; 3:16–21.

33.

Kourtis

, Nesheim

, Thea

, Nahmias

, Lee

. Correlation between soluble L-Selectin, an immune activation marker, and virus replication in pediatric HIV infection. AIDS, 2000; 14:2429–2436.

34.

Fischetti

, Danso

, Dompreh

, Addo

, Haaheim

, Allain

. Vertical transmission of HIV in Ghanaian women diagnosed in cord blood and post-natal samples. J Med Virol, 2005; 77:351–359.

35.

Pascual

, Bruna

, Cerrolaza

et al. Absence of maternal-fetal transmission of HIV-1 to second-trimester fetuses. Am J Obstet Gynecol, 2000; 183:838–842.

36.

Montano

, Russell

, Gilbert

et al. Comparative prediction of perinatal HIV type 1 transmission, using multiple virus load markers. J Infect Dis, 2003; 188:406–414.