Zheng, Junying PhD, MD*; Xie, Yiming MS†; Campbell, Richard PhD‡; Song, Jun PhD, MD*; Wang, Rose Q MD*; Chiu, Robert PhD, MD*; Berenson, James MD‡; Razi, Miriam BS*; Massachi, Samira BS*; Yang, Otto O MD§; Chen, Irvin S Y PhD†; Pang, Shen PhD*
HIV infection of CD4− cells has been reported since 1987. HIV infection of brain cells,1-3 epithelial cells derived from the upper and lower human female reproductive tract,4 primary intestinal epithelial cells,5 cervical epithelial cells,6 mammary epithelial cells, and oral mucosal cells7,8 is important in the understanding of viral latency and transmission.
Epithelial cells covering mucosal surfaces are the cells that initially come in contact with HIV. Thus, these cells play a critical role in early stages of HIV-1 transmission. The mechanism of viral transmission across the epithelium of the genital tract, colon, and oral cavity remains unidentified. HIV may penetrate via lesions in the epithelium.9,10 Lesions are not required for infection, however, which suggests that the virus is capable of crossing the epithelial barrier by other mechanisms.11-13
One theory asserts that HIV can be carried by dendritic cells (DCs) to lymphocytes. Moreover, a type II integral membrane protein, DC-SIGN, which is expressed on the surface of DCs, is capable of binding primary R5 and laboratory-adapted X4 strains of HIV-1 to DCs.14,15 There are several unanswered questions, however. An important inconsistency is that DCs within the epithelium do not express DC-SIGN.16,17 Another dilemma is that there are only low numbers of DCs found on the surface of the epithelium,18-20 with the exception of particular locations, such as intestinal lumen, where a greater number of DCs are found.21 Because DCs are sparse in the epithelium and only a few DCs express DC-SIGN, the total number of DC-SIGN-positive DCs available for HIV attachment is sparse. Therefore, although in cell culture conditions, DC can use its surface lectin, DC-SIGN, to capture HIV particles and release them to CD4+ cells, such a mechanism may not be essential for HIV transmission in vivo.
We hypothesized that HIV can cross the epithelium barrier by infecting epithelial cells. Our previous studies and those conducted by others have demonstrated that HIV infects epithelial cells.4-8,22-26 HIV may complete its transmission process by infecting epithelial cells from the apical side, replicating in the infected cells, and then discharging new viral particles from the basolateral side.
There are a few reports describing HIV transmission through infection of epithelial cells. None of these reports perform quantitative studies of the number or percentage of infected cells, however. These previous studies used indirect methods to demonstrate the infection of HIV. For example, HIV infection of primary uterine epithelial cells has been reported by Asin et al.22 In this specific report, HIV infection was measured by p24 assay. Although p24 assay reveals the replication of HIV, it does not demonstrate the precise percentage of cells infected. Without a precise quantification of HIV infection, it is difficult to assess the significance of HIV infection of epithelial cells in the course of HIV transmission.
Studies of HIV infection of CD4− cells may also bear on the understanding of persistence of HIV infection and latency. Viral infection of neurons and glial cells in the central nervous system (CNS) has been reported.1-3,27-29 Because of the long life span of cells in the CNS, HIV infection of these cells is a probable mechanism of HIV latency and persistent viral infection. Hence, a quantitative study of HIV infection of brain cells is important in understanding the pathogenic effects of HIV in the brain.
To quantify HIV infection of CD4− cells that are involved in HIV transmission and viral latency, we constructed HIV clones carrying the enhanced green fluorescent protein (EGFP) gene.24,30 HIV-infected cells can be quantified by the expression of EGFP in the infected cells.24,26,30 The use of EGFP gene expression virus offers a direct way to quantify HIV infection, unlike the p24 assay, which offers an indirect quantification of HIV infection. Moreover, the number and percentage of infected cells can be easily tabulated using fluorescent activated cell sorting (FACS).
We previously used EGFP-modified HIV clones to study HIV infection of oral epithelial cells and found that HIV infected 2 oral epithelial cell lines, TU177 and TU139, at approximately 1% to 3% when 104 cells/well in 24-well plates were infected with 100 ng of p24 Gag per 0.5 mL.24,26,30 Although the infection rate of primary oral cells was lower, it was still noteworthy.30 These studies suggest that HIV infection of epithelial cells is a plausible mechanism of HIV transmission. In this study, we investigated HIV transmission by using cell lines derived from the female reproductive tract, colon, and intestine. One vaginal (VK2/E6E7) and 2 cervical (GH329 and Ca SKi) cell lines were also used to quantify HIV infection of epithelial cells in the female reproductive tract. So too, 1 colon (CaCo-2) and 1 intestinal (FHS 74 Int) cell line were used to study HIV infection of epithelial cells in the gastrointestinal tract. Furthermore, we examined infectivity of a neuronal cell line (HCN-2) and a glial cell line (A172), because HIV infection of brain cells may be related to AIDS dementia.
It is also necessary to study the way by which HIV infects CD4− cells. One proposal has linked the binding of gp120 glycoprotein to chemokine receptors or galactosylceramide (GalCer) on target cells31-33 to HIV infection. Alternatively, Asin et al22 proposed that the expression of CD4 molecules on certain types of epithelial cells is the mechanism of HIV infection. Both of these hypotheses emphasize that gp120 is essential in viral infection of CD4− cells. Our approach to the study of the role of gp120 in the viral infection of CD4− cells is to use gp120-truncated HIV clones, however.24,26,30 Our previous studies demonstrated that HIV infection of numerous types of epithelial cells is independent of gp120. Within this study, we used gp120-defective HIV to infect epithelial cells from the female reproductive system, colon, and glioneuronal cells. The role of gp120 in viral infection of CD4− cells was clarified.
MATERIALS AND METHODS
Cell Lines and Maintenance
Cell lines A172, HCN-2, GH329, CaCo-2, Ca SKi, VK2/E6E7, LNCaP, and FHS 74 Int were purchased from the American Type Culture Collection (ATCC). The SV40-transformed embryonic kidney cell line (293T) and the HeLa and HeLa-CD4 cell lines have been described previously.24,26 Cell lines LNCaP, 293T, HeLa, HeLa-CD4, and Ca SKi were maintained in RPMI medium supplied with 10% fetal bovine serum (FBS). Cell lines A172, HCN-2, GH329, and FHS 74 Int were maintained in Dulbecco modified Eagle medium (DMEM) medium with 10% FBS. Cell line CaCo-2 was maintained in DMEM medium supplemented with 20% FBS, and VK2/E6E7 was maintained in serum-free keratinocyte growth medium (KGM; Invitrogen).
The oral cell line, human oral tissue (HOT), was derived from oral epithelial cells from a patient. The oral epithelial cells derived from the patient were maintained in KGM with 10% FBS for a few passages. A colony of cells in a plate indicating a high growth rate was selected and grown in RPMI medium with 10% FBS. After 5 passages, the cell culture was stabilized as a cell line. The cell line showed a growth rate similar to the 2 previously reported oral cell lines, TU139 and TU177, respectively. Human peripheral blood leukocytes (PBLs) were obtained from the UCLA AIDS Viral Core Laboratory. These cells were stimulated by phytohemagglutinin (PHA; 1 μg/mL) and interleukin-2 (IL-2).
HIV Virus With Enhanced Green Fluorescent Protein Gene Expression
The HIVNL4-3-EGFP-Env(+) virus is derived from HIVNL4-3, with part of the nef gene sequence replaced by the EGFP gene. The HIVNL4-3-EGFP-Env(−) virus is structurally similar to HIVNL4-3-EGFP-Env(+), except for a sequence deletion between the 2 Bgl II sites (nt 7032-7612).24 Deletion of 581 nucleotides truncates the C-terminal 230-amino acid residues of gp120, and thus introduces a frameshift in gp160 downstream from the deletion site. Therefore, the resulting gp160 mRNA can no longer serve as a template for the generation of gp41.
Using a similar approach, we constructed an EGFP gene-modified clade B, R5 strain, JRCSF34, and a clade D, R5 strain, 94UG114.35
DsRed Expression of HIVNL4-3
A similar method to that described previously was used to insert the DsRed gene (BD Clontech, Mountain View, CA) into a plasmid that contains the gp120 truncated HIVNL4-3 genome. The resulting plasmid was used to cotransfect 293T cells with the helper vector pLET, which carries HIV Env gene derived from HIVLAI.26 The generated virus was titrated by infecting HeLa-CD4 cells by limited dilution. Infected cells were detected by their red color under a fluorescent microscope.
We prepared HIVNL4-3-EGFP-Env(+) from 293T cells by transfection.24,26 To remove cellular debris, the collected supernatant derived from the transfected 293T cells was filtered by passing it through a 0.2-μm filter. HIVNL4-3-EGFP-Env(−) and the DsRed expression HIVNL4-3 were prepared from 293T cells by an analogous method.
We also prepared HIVNL4-3-EGFP-Env(+) from PBLs, CEM (a T lymphocyte cell line derived from patient CEM), HOT (oral cell line), GH327, A172, and Ca SKi. Because of low transfection efficiencies in these cell lines, we used an infection procedure that introduces the viral genome into cells.24,26 PBLs and CEM were infected by HIVNL4-3-EGFP-Env(+) generated from 293T cells. Three hours after infection, the cells were washed 3 times to remove unbound virus. Culture medium containing virus generated from PBLs or CEM was collected from days 4 to 9 after infection on a daily basis.
The HIVNL4-3-EGFP-Env(+) virus was titrated by limited dilutions of viral stocks to infect CEM CD4+ cells per well in 96-well plates. The most diluted viral dosage capable of infecting these cells was designated as 1 tissue culture infectious dose (TCID). As many previous studies suggest, we also used a p24 assay to titer HIV. HIVNL4-3-EGFP-Env(−) was generated from 293T or GH329 cells. Because the Env(−) virus was not able to infect HeLa-CD4 or CEM cells, we used the p24 count to indicate viral titer.
Infection of Epithelial and Brain Cells
We plated 2 × 104 each of A172, HCN-2, GH329, CaCo-2, Ca SKi, VK2/E6E7, and FHS 74 Int cells into 24-well plates 24 hours before infection. A viral aliquot with 105 TCID [or 100 ng of p24 in the case of using HIVNL4-3-EGFP-Env(−)] was added to the cell cultures at a 0.5-mL total volume. To avoid system errors, all infections were duplicated. We presumed that the cell numbers doubled during the 24-hour period. Hence, the concentration of virus was 105 viral TCID/4 × 104 cells in 0.5 mL for most cell lines and 105 viral TCID/2 × 104 for LNCaP. At 16 hours after infection, the cell cultures were washed 4 times to remove the unbound virus.
Quantification of Enhanced Green Fluorescent Protein-Positive Cells
The expression of EGFP was visualized and counted using fluorescent microscopy or FACS.24
Infection of Cells in a Transwell Device
Methods to use cell cultures in a transwell device have been described by Collins et al.36 Cells (104) were plated in the top chambers in 12-well transwell culture plates. The top chambers are separated from the bottom chambers by a membrane with 3.0-μm pores. When the cells were confluent, HIV (100 ng of p24) was used to infect the cell monolayer in the top chambers. After infection, the cells were washed 4 times to remove unbound viral particles. Two days after infection, the top chambers were moved to 12-well plates with HeLa-CD4 cells plated in the bottom chambers. The release of virus generated from the basolateral side of the cell monolayer was quantified by counting EGFP-positive HeLa-CD4 cells in the bottom chambers.
To estimate paracellular leakage of the cell monolayers, we used a DsRed gene-modified HIVNL4-3. At 2 days after infection, the EGFP-HIVNL4-3-infected cell cultures in the top chambers were moved to fresh plates with HeLa-CD4 cells in the bottom chambers. At this time, an aliquot (100-ng p24 count) of DsRed expression HIVNL4-3 was added to the top chambers. Top chambers containing no cells were used as the controls. Two days after the addition of DsRed expression HIV to cell monolayers, the DsRed HIV-infected cells in the bottom chambers were counted. The paracellular leakage of cell monolayers was calculated by comparing the numbers of DsRed-positive cells in the wells, which contained cell monolayers in the top chambers, with the control wells, which contained no cells in the top chambers.
Polymerase Chain Reaction Analysis of Viral DNA
Infected cells in wells of 24-well plates were collected, and the DNA from these cells was isolated. The isolated DNA samples were subjected to polymerase chain reaction (PCR) analysis using the PCR primer pair M661/M66737 or the PCR primer pair M667/AA55.38 The PCR primer pair M661/M667 detected full-length viral DNA, whereas the PCR primer pair M667/AA55 detected the long terminal repeat (LTR) sequence in partially and completely reverse-transcribed viral DNA.
p24 Assays of Infected Cell Cultures
After infection, cells were washed 4 times to remove unbound virus. This was followed by the addition of 0.5 mL of growth medium into each well of the cell cultures. The culture medium was changed daily, and medium was collected at days 0, 2, 4, 6, and 8 after infection. Subsequently, p24 assays were performed.
Infection of Cells Derived From the Female Reproductive Tract, Colon, Intestine, and Brain With Virus Prepared From 293T Cells
We infected various cell lines with HIVNL4-3-EGFP-Env(+) virus prepared from 293T cells (Fig. 1A). The CD4+ T-lymphocyte cell line, CEM, and HeLa cells were used as positive and negative controls, respectively. The HeLa cell line was used as a negative control, because previous studies demonstrated that HIV does not infect this cell line.24,26 Our results indicated that HIV infected VK2/E6E7, A172, GH329, and Ca SKi cells at rates of 1.3% to 3.1% (Fig. 1B). Infection rates of Caco-2 and HCN-2 cells were lower but still indicated an easily detectable level of infection, which was approximately 0.3% to 0.6% (Fig. 1B).
We studied whether low doses could infect CD4− cells and whether HIV could infect confluent cell cultures. We found that virus with a concentration of 5 × 103 TCID/mL was able to infect A172, GH329, and Ca SKi cells. Subconfluent and confluent cells could be infected; yet, we found that the same amount of virus infected more cell colonies in the confluent cell cultures (Fig. 1C). This finding might be attributable to the fact that there were more cells available for infection during the time virus was added into these cell cultures.
To investigate infection kinetics, we used azidothymidine (AZT) to inhibit reverse transcription at time points of 0, 2, 6, 24, and 48 hours after infection. Although the addition of AZT showed significant inhibition 0 or 2 hours after infection, it resulted in only moderate inhibition when it was added 6 hours after infection. The profiles of AZT inhibition of HIV infection of CD4− cell lines and HeLa-CD4 (a CD4+ cell line) were similar (Fig. 2A), which suggests that in CD4− cells, HIV can also complete reverse transcription in less than 24 hours. To confirm that the AZT inhibition was directly caused by inhibiting reverse transcription, we quantified the proviral DNA in the infected cells. To do so, we collected the cells treated with AZT at different time points and used PCR analysis to quantify the viral DNA levels. The PCR primer pair M667/AA5538 was used to quantify viral DNA containing LTR sequences, and the M667/M66137 primer pair was used to quantify full-length reverse-transcribed viral DNA. Our results demonstrated that AZT effectively inhibited the reverse transcription of the full-length viral DNA (Fig. 2B), which suggests that reverse transcription is required for the viral infection of CD4− cells.
We also examined reverse transcription by PCR analysis of viral DNA at different time points with no AZT added. The CD4+ cell line, CEM, and the CD4− cell line, A172, were infected by HIV, and DNA was isolated from infected cells 2, 6, 24, 48, and 120 hours after infection. Comparing the viral DNA profiles of HIVNL4-3-EGFP-Env(+)-infected CEM and A172 cells, we found that HIV reverse transcription in these 2 cell lines was similar. The levels of full-length viral DNA were low for both cell lines 2 hours after infection. At 6 hours, some full-length viral DNA was detected in both infected cell lines. The levels of full-length viral DNA in the samples isolated 24, 48, and 120 hours after infection were almost identical, suggesting that reverse transcription was completed 24 hours after infection (Fig. 2C). These results were consistent with the results of AZT inhibition studies shown in Figures 2A, B.
Virus Replication in Infected Cells
To determine whether HIV could replicate in infected epithelial and brain cells and generate infectious progeny virus, we collected the culture media from the infected cell cultures and used them to infect CD4+ cells, HeLa-CD4 cells, and CEM. Our results demonstrated that the cell culture medium samples collected from infected A172, GH329, Ca SKi, VK2/E6E7, and Caco-2 cultures were infectious for HeLa-CD4 (Fig. 3A) and CEM cells. Culture medium collected from the negative control (HeLa cells) did not show infection, which suggests that infectious virions could only be generated from infected cells. We also measured the levels of p24 Gag protein in the collected medium during the time course. The results demonstrated that p24 counts increased along the time course, which indicates that the quantity of the viral particles released from the infected cells increased (Fig. 3B). These results demonstrated that HIV could infect epithelial and brain cells and replicate and generate infectious progeny virus. By combining the numbers of HIV infection of CD4− cells, the numbers of HeLa-CD4 cells infected by virus generated from the infected CD4− cells, and the p24 counts as shown in Figure 3B, we were able to determine the numbers of infectious viral particles per infected CD4− cells (Fig. 3C). Our results demonstrated that each infected CD4− cell generated 0.6 to 13 infectious virions that can infect HeLa-CD4 cells (Fig. 3C). According to our titration studies, 1 HeLa-CD4 TCID can infect approximately 5 CEM cells. Therefore, we estimated that each infected CD4− cell could generate 3 (5 × 0.6) to 65 (5 × 13) TCIDs of infectious virions titrated by CEM cells.
Discharge of Viral Particles From the Basolateral Side of the CD4− Cell Monolayers
Transwell devices were used to study the amount of virions discharged from the basolateral sides of cell monolayers. To quantify these amounts, we infected cervical cell lines GH329 and Ca SKi in the top chambers of transwell plates. In addition, the viral-nonsusceptible cell line HeLa was used as a negative control. After infection, we washed the infected cell cultures 4 times to remove all unbound viral particles. Two days after infection, EGFP-positive cells were observed under a fluorescent microscope, suggesting that some cells were infected by NL4-3-EGFP-Env(+) HIV virus. At this point, we transferred the top chambers that contained the infected cells to new 12-well plates with 4 × 104 HeLa-CD4 cells in each well. We expected that some of the viral particles released from the basolateral side of the infected cell monolayer could pass through the 3.0 pores of the transwell filter. Because NL4-3-EGFP-Env(+) HIV is infectious for HeLa-CD4, the released virus could be quantified by counting the EGFP-positive HeLa-CD4 cells.
It is important to estimate the paracellular leakage of cell monolayers, because virus released from apical sides could use paracellular leakage to pass through cell monolayers and then pass the transwell membrane to infect HeLa-CD4 cells in the bottom chambers. Therefore, we used DsRed gene expression HIVNL4-3 to estimate the levels of paracellular leakage that occur in a cell monolayer. Much like the EGFP expression HIVNL4-3, the DsRed expression HIV infects HeLa-CD4 cells with high efficiency, whereas it infects CD4− cells with lower efficiency. Hence, it is expected that the viral particles passing through the monolayers and the transwell membranes were able to infect HeLa-CD4 cells in the bottom chamber. We applied DsRed expression HIVNL4-3 virus with a 100-ng p24 count to top chambers that did not contain any cells. After 16 hours, the top chambers were removed from the transwell plates. The DsRed expression viral particles that could pass through the monolayer of epithelial cells and the transwell membrane during this period of 16 hours were expected to infect the HeLa-CD4 cells in the bottom chamber. At 3 days after infection, the DsRed gene expression HIV-infected HeLa-CD4 cells were counted. Our results demonstrated that there were low percentages of viral particles able to pass through confluent cell monolayers (Table 1). The monolayers of GH329 cells allowed only 4.1% and the monolayers of Ca SKi cells allowed only 1.4% passage of virions to pass through the monolayers. Therefore, there was only a low percentage of virus released from the apical side of monolayers that could pass through the monolayers and infect HeLa-CD4 in the bottom chamber. In other words, it is safe to assume that most EGFP expression HeLa-CD4 cells in the bottom chambers were infected by virus released from the basolateral sides.
We found that infectious viral particles were released from almost all the tested cell lines except the negative control, HeLa (Table 1). These results suggest that the virus can enter epithelial cells from the apical side and the replicated virus can bud out from the basolateral side. We noted that the numbers of virus passing through the transwell membrane were approximately 10-fold lower than the total number of virus released from the infected cells (comparing Fig. 3A with Table 1). One explanation for this is that only a small percentage of viral particles released from the basolateral side can pass through the pores in the transwell membrane. To quantify the percentage of virus that could pass through the pores of the transwell membrane, we set up 2 different procedures for infection. The first method of infection was performed by adding virus directly into HeLa-CD4 culture, and the second method of infection was performed by adding virus to the top chambers of the transwell plates. We found that the transwell membrane only allows 10% of HIV particles to pass through (Table 1). Thus, it is estimated that only approximately 10% of total viral particles released from the basolateral side of the infected cells were able to pass through the pores.
Infection of Epithelial and Glial Cell Lines by Virus Prepared From Peripheral Blood Leukocytes, CEM T-Lymphocytes, GH329 and Ca SKi Cervical Cells, and HOT Oral Cells
Our results indicate that HIV can be generated from certain CD4− cells. Therefore, in HIV-infected individuals, HIV can arise from several types of cells, including infected CD4+ T lymphocytes, macrophages, and, possibly, CD4− cells. Determining the infectivity of HIV from different cell types in infected epithelial cells enables us to assess the importance of HIV infection of CD4− cells during disease transmission.
We prepared HIV from PBLs and 6 cell lines, CEM, GH329, Ca SKi, FHS 74 Int, A172, and HOT. We selected these cell lines, because virus generated from T lymphocytes and epithelial cells from the female genital tract or oral cavity is expected to be directly related to viral transmission. The viral preparations were used to infect A172, Ca SKi, GH329, and VK2/E6E7 cells. Virus generated from PBLs, CEM, GH329, and HOT cells demonstrated significant infectivity for these 4 CD4− cell lines: A172, Ca SKi, GH329, and VK2/E6E7. Virus generated from the A172, FHS 74 Int, and Ca SKi cell lines demonstrated much lower infectivity for the tested epithelial and brain cell lines, however (Fig. 4). These results suggest that virus generated from T lymphocytes and certain types of epithelial cells can potentially infect various CD4− cells. The susceptibility of the glial cell line A172 to infection by HIV generated from several types of cells suggests that glial cells may serve as latent hosts for HIV persistent infection. It is also notable that virus generated from PBLs is not as infectious as virus generated from GH329, HOT, and CEM cells. It is possible that PBLs are a mixture of CD4+, CD8+, and B cells as well as other leukocytes. Virus generated from CD4+ T lymphocytes may infect CD4− cells with similar efficiency as virus generated from the CD4+ cell line, CEM. The virus generated from other types of leukocytes may not be as infectious as CD4+ T cells in infection of A172, GH329, or Ca SKi cell lines, however.
Infection of Epithelial and Glial Cell Lines Through a gp120-Independent Mechanism
Whether gp120 is a critical factor in HIV infection of CD4− cells is controversial. To determine its roles, we used HIV clones with intact or truncated gp120 to infect CD4− cells and then compared the infectivity. We generated a gp120-truncated HIV clone and used Western blot analysis to confirm that the prepared EGFP-modified HIVNL4-3 Env(−) virus had neither gp120 nor gp160 (Fig. 5A). Also, we assumed that the gp41 glycoprotein was absent in such a viral clone because of the absence of gp160, a precursor of gp41. EGFP-modified HIVNL4-3 with intact or truncated gp120 prepared from GH329 and 293T cells was used to infect Caco-2, VK2/E6E7, GH329, Ca SKi, LNCaP, and A172 cells. The infection rates of these 2 viral clones in these 6 cell lines were almost identical, regardless of whether the gp120 was truncated or intact (Figs. 5B, C), suggesting that the role of gp120 is minimal.
So too, in previous studies, chemokine receptors CXCR4 and CCR5 were believed to be key proteins involved in HIV infection of CD4− cells. To study the roles of these 2 membrane proteins, we transfected GH329, Ca SKi, and 293T cell lines with plasmids that carried the CXCR4 or CCR5 gene. We used a plasmid with the firefly luciferase gene driven by the same promoter, the cytomegalovirus (CMV) promoter, as a control. Compared with the control cell cultures that were transfected with the luciferase-expressing plasmid, the CXCR4 or CCR5 gene-transfected cells demonstrate no difference in viral infection (Fig. 5D). Once again, these results suggest that neither CXCR4 nor CCR5 was a protein necessary for HIV infection of CD4− cells.
Infection of Epithelial and Brain Cells by HIV Strains JRCSF and 94UG114
Our results demonstrated that HIVNL4-3, an X4 viral clone in clade B, infected epithelial and glial-neuronal CD4− cells. In addition to such findings, we attempted to determine whether R5 viral strains from clade B or other clades could infect epithelial and brain cells. In particular, we used a clade D and R5 virus strain, 94UG114, and a clade B/R5 strain, JRCSF, to infect CD4− cells. We prepared EGFP HIVJRCSF and HIV94UG114 from 293T cells by transfection. The viral preparations were used to infect A172, VK2/E6E7, GH329, Ca SKi, and HeLa cells. The infection profiles were similar to the infection of these CD4− cells by EGFP expression HIVNL4-3, as shown in Figure 1B, except that the infectivities of these 2 viral clones were approximately 3- to 4-fold lower than the infectivity of HIVNL4-3 for cell lines GH329, A172, Ca SKi, and VK2/E6E7. In Figure 6, a gp120-defective JRCSF strain demonstrated infectivity similar to that for 94UG114. These results suggest that X4 and R5 viral strains from various subtypes can infect epithelial and brain cells through gp120-independent mechanisms.
The mechanism by which HIV crosses the epithelium during viral transmission is not identified. Although previous reports have suggested that HIV may infect epithelial cells to cross the epithelium, few quantitative studies have been performed to prove the validity of this assumption. So too, the mechanism of HIV infection of brain cells, including glial and neuronal cells, is unknown. Because brain cells have a long life span, infection of these cells may serve as an HIV reservoir. Therefore, HIV infection of CD4− cells, including epithelial and brain cells, can be an important mechanism of viral transmission and persistent infection.
The use of an EGFP-modified viral strain provides a direct and sensitive approach to quantify HIV infection. We infected 5 epithelial cell lines derived from the female reproductive or gastrointestinal tract and 2 glioneuronal cell lines from the brain. Our results demonstrated that HIV infected all these cell lines, although some had lower susceptibility to infection. Viral replication was also demonstrated, and the progeny virus generated from these infected cells was infectious. These results, combined with our previous studies of oral epithelial cells,30 strongly suggest that HIV can infect CD4− cells and use this mechanism to traverse the epithelium during transmission and to generate an HIV brain reservoir for persistent infection.
Although our results demonstrated that HIV infects various epithelial cells, the physiologic significance of such infection requires additional studies. For example, vaginal epithelium is covered with nonkeratinized stratified squamous epithelial cells. HIV may not be able to cross such a multilayer of cells before infecting CD4+ lymphocytes. It is expected that HIV infection of epithelial cells that cover the surface of the gastrointestinal tract or cervix may be more important, because the cavity of these tissues is covered by a layer of simple columnar epithelial cells. Should these cells become infected, virus released from the basolateral sides of such cells would easily come in contact with CD4+ cells.
Some studies suggest that HIV infects epithelial and brain cells through the binding of HIV gp120 to chemokine receptors or GalCer on target cells, thereby inferring that gp120 plays an essential role in CD4− cell infection.36-38 Previous reports also suggested that for some types of epithelial cells, the expression of CD4 molecules on the surface of these cells may be the crucial factor in infection.22 In other words, such research conjectures that gp120 is essential for the infection of epithelial and brain cells. Our previous studies24,26 and the results shown in this report demonstrate that gp120-defective HIV has infectivity similar to HIV with intact gp120, however. In addition to data from our previous studies, we have tested 14 different epithelial or brain cell lines and oral primary cells, none of which demonstrated a significant difference when infected with intact or defective gp120. Because the studies by others used different viral strains and cell lines than we utilized in our research, we cannot exclude the possibility that some HIV strains use gp120 to bind to chemokine receptors or GalCer for viral infection. Undeniably, however, gp120-independent infection is a significant mechanism by which HIV infects CD4− cells.
Our results demonstrate that some cell lines are more susceptible to HIV infection than others. The 3 cell lines derived from the female reproductive tract, VK2/E6E7, GH329, and Ca SKi, and 1 glial cell line, A172, demonstrated much higher susceptibility (1.3%-3.1% infection by virus at a concentration of 105 TCID in 0.5 mL/well) than the control cell line, HeLa, and the other 3 cell lines, Caco-2, HCN-2, and FHS 74 Int. One explanation for this phenomenon is that the VK2/E6E7, GH329, Ca SKi, and A172 cell lines express specific membrane proteins at higher levels than the other 3 cell lines. Furthermore, the control cell line, HeLa, may not express these membrane proteins at all.
The authors thank Wendy Aft for editing.
1. Cheng-Mayer C, Rutka JT, Rosenblum ML, et al. Human immunodeficiency virus can productively infect cultured human glial cells. Proc Natl Acad Sci USA. 1987;84:3526-3530.
2. Kunsch C, Hartle HT, Wigdahl B. Infection of human fetal dorsal root ganglion glial cells with human immunodeficiency virus type 1 involves an entry mechanism independent of the CD4 T4A epitope. J Virol. 1989;63:5054-5061.
3. Ensoli F, Ensoli B, Thiele CJ. HIV-1 gene expression and replication in neuronal and glial cell lines with immature phenotype: effects of nerve growth factor. Virology. 1994;200:668-676.
4. Howell AL, Edkins RD, Rier SE, et al. Human immunodeficiency virus type 1 infection of cells and tissues from the upper and lower human female reproductive tract. J Virol. 1997;71:3498-3506.
5. Chenine AL, Matouskova E, Sanchez G, et al. Primary intestinal epithelial cells can be infected with laboratory-adapted strain HIV type 1 NDK but not with clinical primary isolates. AIDS Res Hum Retroviruses. 1998;14:1235-1238.
6. Tan X, Phillips DM. Cell-mediated infection of cervix derived epithelial cells with primary isolates of human immunodeficiency virus. Arch Virol. 1996;141:1177-1189.
7. Qureshi MN, Barr CE, Hewlitt I, et al. Detection of HIV in oral mucosal cells. Oral Dis. 1997;3 (Suppl 1):S73-S78.
8. Toniolo A, Serra C, Conaldi PG, et al. Productive HIV-1 infection of normal human mammary epithelial cells. AIDS. 1995;9:859-866.
9. Dickerson MC, Johnston J, Delea TE, et al. The causal role for genital ulcer disease as risk factor for transmission of human immunodeficiency virus. Sex Transm Dis. 1996;23:429-440.
10. Kozak SL, Platt EJ, Madani N, et al. CD4, CXCR-4, and CCR-5 dependencies for infections by primary patient and laboratory-adapted isolates of human immunodeficiency virus type 1. J Virol. 1997;71:873-882.
11. Miller CJ. Animal models of viral sexually transmitted diseases. Am J Reprod Immunol. 1994;31:52-63.
12. Miller CJ, Alexander NJ, Sutjipto S, et al. Effects of virus dose and nonoxynol-9 on the genital transmission of SIV in rhesus macaques. J Med Primatol. 1990;19:401-409.
13. Spira AI, Marx PA, Patterson BK, et al. Cellular targets of infection and route of viral dissemination after an intravaginal inoculation of simian immunodeficiency virus into rhesus macaques. J Exp Med. 1996;183:215-225.
14. Geijtenbeek TB, Kwon DS, Torensma R, et al. DC-SIGN, a dendritic cell-specific HIV-1-binding protein that enhances trans-infection of T cells. Cell. 2000;100:587-597.
15. Pohlmann S, Soilleux EJ, Baribaud F, et al. DC-SIGNR, a DC-SIGN homologue expressed in endothelial cells, binds to human and simian immunodeficiency viruses and activates infection in trans. Proc Natl Acad Sci USA. 2001;98:2670-2675.
16. Turville SG, Arthos J, Donald KM, et al. HIV gp120 receptors on human dendritic cells. Blood. 2001;98:2482-2488.
17. Turville SG, Cameron PU, Handley A, et al. Diversity of receptors binding HIV on dendritic cell subsets. Nat Immunol. 2002;3:975-983.
18. Sobhani I, Walker F, Aparicio T, et al. Effect of anal epidermoid cancer-related viruses on the dendritic (Langerhans') cells of the human anal mucosa. Clin Cancer Res. 2002;8:2862-2869.
19. Giannini SL, Hubert P, Doyen J, et al. Influence of the mucosal epithelium microenvironment on Langerhans cells: implications for the development of squamous intraepithelial lesions of the cervix. Int J Cancer. 2002;97:654-659.
20. Zavala WD, De Simone DS, Sacerdote FL, et al. Variation in Langerhans cell number and morphology between the upper and lower regions of the human esophageal epithelium. Anat Rec. 2002;268:360-364.
21. Niess JH, Brand S, Gu X, et al. CX3CR1-mediated dendritic cell access to the intestinal lumen and bacterial clearance. Science. 2005;307:254-258.
22. Asin SN, Fanger MW, Wildt-Perinic D, et al. Transmission of HIV-1 by primary human uterine epithelial cells and stromal fibroblasts. J Infect Dis. 2004;190:236-245.
23. Liu X, Zha J, Chen H, et al. Human immunodeficiency virus type 1 infection and replication in normal human oral keratinocytes. J Virol. 2003;77:3470-3476.
24. Pang S, Yu D, An DS, et al. Human immunodeficiency virus Env-independent infection of human CD4(−) cells. J Virol. 2000;74:10994-11000.
25. An DS, Xie Y, Chen IS. Envelope gene of the human endogenous retrovirus HERV-W encodes a functional retrovirus envelope. J Virol. 2001;75:3488-3489.
26. Chow YH, Yu D, Zheng JY, et al. gp120-Independent infection of CD4− epithelial cells and CD4+ T-cells by HIV-1. J Acquir Immune Defic Syndr. 2002;30:1-8.
27. An SF, Groves M, Giometto B, et al. Detection and localisation of HIV-1 DNA and RNA in fixed adult AIDS brain by polymerase chain reaction/in situ hybridisation technique. Acta Neuropathol (Berl). 1999;98:481-487.
28. Schweighardt B, Shieh JT, Atwood WJ. CD4/CXCR4-independent infection of human astrocytes by a T-tropic strain of HIV-1. J Neurovirol. 2001;7:155-162.
29. Lawrence DM, Durham LC, Schwartz L, et al. Human immunodeficiency virus type 1 infection of human brain-derived progenitor cells. J Virol. 2004;78:7319-7328.
30. Zheng JY, Yang OO, Xie Y, et al. Ethanol stimulation of HIV infection of oral epithelial cells. J Acquir Immune Defic Syndr. 2004;37:1445-1453.
31. Harouse JM, Bhat S, Spitalnik SL, et al. Inhibition of entry of HIV-1 in neural cell lines by antibodies against galactosyl ceramide. Science. 1991;253:320-323.
32. Yahi N, Sabatier JM, Nickel P, et al. Suramin inhibits binding of the V3 region of HIV-1 envelope glycoprotein gp120 to galactosylceramide, the receptor for HIV-1 gp120 on human colon epithelial cells. J Biol Chem. 1994;269:24349-24353.
33. Fantini J, Hammache D, Delezay O, et al. Sulfatide inhibits HIV-1 entry into CD4−/CXCR4+ cells. Virology. 1998;246:211-220.
34. Koyanagi Y, Miles S, Mitsuyasu RT, et al. Dual infection of the central nervous system by AIDS viruses with distinct cellular tropisms. Science. 1987;236:819-822.
35. Gao F, Robertson DL, Carruthers CD, et al. A comprehensive panel of near-full-length clones and reference sequences for non-subtype B isolates of human immunodeficiency virus type 1. J Virol. 1998;72:5680-5698.
36. Collins KB, Patterson BK, Naus GJ, et al. Development of an in vitro organ culture model to study transmission of HIV-1 in the female genital tract. Nat Med. 2000;6:475-479.
37. Pang S, Koyanagi Y, Miles S, et al. High level of unintegrated HIV-1 DNA in brain tissue of AIDS dementia patients. Nature. 1990;343:85-89.
38. Zack JA, Arrigo SJ, Weitsman SR, et al. HIV-1 entry into quiescent primary lymphocytes: molecular analysis reveals a labile, latent viral structure. Cell. 1990;61:213-222.
© 2006 Lippincott Williams & Wilkins, Inc.