Leukocyte infiltrates that can serve as viral reservoirs, and sites for viral replication are found in many organs of HIV-1-infected patients. Patients whose blood leukocytes migrate across confluent endothelial monolayers ex vivo and transmit infectious virus to mononuclear leukocytes (MNLs) lodged beneath this endothelial barrier have a worse prognosis. We evaluated the ability of 110- to 120-kDa fibronectin fragments (FNf), which are found in the blood of >60% of HIV-1-infected patients, to stimulate transendothelial migration and drive productively infected MNLs into a potential perivascular space. FNf induced MNLs to release TNF-α in a dose-dependent fashion; the resulting increase in lymphocyte and monocyte transendothelial migration could be blocked with soluble TNF receptor I. Rather than penetrate deeply into the subendothelial matrix, as is seen with untreated controls, FNf-treated MNLs clustered just below the endothelial monolayer. Treatment with FNf during migration increased subsequent recovery of HIV-infected cells from the subendothelial compartment. FNf treatment also significantly increased the numbers of HLA-DRbright, dendritic-type cells that reverse-migrated from the subendothelial depot to the apical endothelial surface 48 h after migration. Fibronectin fragments can be produced by viral and host proteases in the course of inflammatory conditions. The ability of FNf to stimulate transendothelial migration of HIV-1-infected MNLs may help to explain the dissemination of this infection into cardiac, renal, and CNS tissues.

Human immunodeficiency virus type 1 infections are associated with disorders not only of central and peripheral nervous systems (1, 2), but also of heart (3, 4), skeletal muscle (5, 6), and other organs (7, 8, 9). A common histopathologic feature of all these extralymphoid manifestations of HIV-1 infections is the accumulation of activated monocytes and lymphocytes in multiple perivascular foci distributed at random in the vasculature (5, 10, 11). Frequently, these foci incorporate virus-infected leukocytes that may be a source of toxic viral products (12), proinflammatory cytokines (13, 14), reactive oxygen species (15), eicosanoids (4), and other molecules that can kill or disable adjacent parenchymal cells (16). Modeling studies suggest that these leukocyte infiltrates could also provide a protected reservoir for recruitment of uninfected leukocytes, a site for replication of virus, and a source of newly infected lymphocytes and dendritic cells that can migrate to lymph nodes (17). Compared with lymphocytes from healthy donors, lymphocytes from HIV-1-infected subjects are significantly more likely to migrate across confluent endothelial monolayers in vitro (18, 19). A critical step in the evolution of HIV-1 infections appears to be a time when circulating, productively infected lymphocytes acquire the ability to transmigrate and export infectious virus across vascular barriers. In a recent 2.5-year prospective study, patients whose leukocytes displayed this capacity were significantly more likely to develop uncontrollable viremia and secondary infections and to die of their HIV-1 infections (17). This behavior of blood mononuclear leukocytes (MNLs),3 seen in slightly >40% of unselected HIV-1-infected men, did not correlate with the plasma viral load, the frequency of circulating provirus-positive cells, or blood CD4 T cell counts. It was seen even in patients who remained closely adherent to prescribed antiretroviral drug programs.

Although HIV-1 infection stimulates the expression of β integrins that lymphocytes use to adhere to endothelium and fibronectin (FN)-rich tissue matrices (20, 21, 22), the stimuli that induce preferential transendothelial migration of infected cells remain poorly understood. Recent studies have called attention to the possibility that proteolytic fragments or polymers of FN may be among the factors that influence leukocyte behavior in inflamed tissues (23, 24, 25). Certain FN fragments can stabilize and enhance the infectivity of cell-associated virus (24, 25); other FN subunits can activate monocytes and cause them to release TNF-α through a protein kinase C-dependent signaling pathway (26, 27, 28, 29, 30). Because the release of monocyte intracytoplasmic TNF-α drives lymphocyte transendothelial migration (18, 31), we investigated the abilities of FN fragments to stimulate lymphocyte transendothelial migration and create subendothelial foci of infected cells that could disseminate virus-infected cells to distant sites.

Peripheral blood MNLs, prepared by Ficoll/Hypaque density centrifugation, were added to the apical surfaces of confluent monolayers of HUVECs (gift from Dr. C. W. Smith, Baylor College of Medicine, Houston TX). HUVECs, isolated weekly (32), were used at the first or second passage after an ∼3-fold expansion in serum- and growth factor-containing endothelial cell medium (proprietary recipe; Cell Systems, Kirkland, WA). Endothelial monolayers were grown on a matrix of hydrated collagen (Vitrogen; Celtrix, Palo Alto, CA) containing 20 μg/ml native FN (Sigma-Aldrich, St. Louis, MO) (19). MNLs, suspended in RPMI 1640 (Invitrogen Life Technologies, Grand Island NY) with 10% FCS (HyClone, Logan UT) were allowed to migrate, typically for 2 h, through the endothelial monolayer into the collagen matrix that supported the endothelial cells. To assess how deeply migrating MNLs penetrated the collagen gel beneath the HUVECs, we used an inverted microscope to count the number of MNLs per high power field in serial planes, spaced at 100-μm intervals through the collagen pad. Three series of planes were counted for each well.

To enumerate the migratory population, nonadherent cells were washed away with Ca2+- and Mg2+-free HBSS, and cells that were adherent, but had not migrated through the endothelium, were then removed with trypsin (19). Migratory MNLs were subsequently harvested by digesting the subendothelial matrix with collagenase (Sigma-Aldrich). To enumerate the original and migratory populations, cells were incubated with fluorochrome-labeled lymphocyte-specific (CD45) and monocyte-specific (CD14) mAbs (LeucoGATE; Coulter, Hialeah, FL) and were counted by flow cytometry using FlowQuant beads (Coulter) as an internal reference.

To assess the ability of MNLs that had migrated below the endothelial barrier to reverse migrate back from the basal to the apical surface, we allowed MNLs to migrate for the customary 2 h. After the nonmigratory cells were washed away with complete medium, we refed the cultures with medium 199 (Invitrogen Life Technologies) containing 20% human type AB serum. Two days later, the reverse-migrating MNLs were collected by gently washing the surface of the HUVECS with warm complete medium so as not to disturb the integrity of the endothelium.

The media, cytokines, and Abs used in these experiments contained <0.03 endotoxin unit/ml endotoxin as determined by the Limulus amebocyte assay (Associates of Cape Cod, Woods Hole, MA). Purified human FN enriched for the 110- to 120-kDa chymotrypsin fragment, hereafter called FNf, was obtained from Invitrogen Life Technologies (Gaithersburg, MD) or Upstate Biotechnology (Lake Placid, NY) and TNF-α was obtained from R&D Systems (Minneapolis, MN). Controls for the FNf included intact FN (Chemicon International, Temecula, CA) and the connecting segment (CS-1) (33) fragment of FN (gift from R. Baughn, Baylor College of Medicine, Houston, TX).

To evaluate its effects on transendothelial migration, FNf was mixed with MNLs on the apical surface of the HUVECs. This approach to treatment with FNf was critical, because monocytes become highly adhesive within minutes of exposure to FNf. If pretreated in a polypropylene test tube, many monocytes will adhere to the walls of the tube and be lost from the assay. Because of the adhesiveness of activated monocytes, we used siliconized tubes (Sigmacote; Sigma-Aldrich) for cell purification and flow cytometric analyses. In addition, we added 0.1% gelatin (Sigma-Aldrich) to the Dulbecco’s PBS wash solutions after Ab staining to further reduce the loss of monocytes from nonspecific adhesion.

In some experiments TNF-α was mixed with the MNLs as they were added to the endothelial surface. Monocytes exposed to TNF-α are not as adhesive as those exposed to FNf. This allowed us to test the effects of pulse treatment with TNF-α to discriminate between its direct effects on MNLs and those on HUVECs. In pulse treatment experiments, TNF-α (typically at 10 ng/ml) was added to MNLs suspended at 3–5 × 106/ml in RPMI 1640 with 10% FCS at 37°C for the indicated times. Sham-treated MNLs were exposed to medium alone. Pulse-treated MNLs were washed with a 10-fold and then a 100-fold excess of HBSS without Ca2+ or Mg2+ and adjusted to 1 × 106/ml in RPMI 1640 with 10% FCS for the migration assay.

We blocked the effect of TNF-α by adding 0.64 μg/ml soluble TNF-α receptor I (TNF sRI/Fc chimera, TNFRSF1A; R&D Systems) to the MNLs and HUVECs before the addition of FNf or TNF-α. This quantity of TNF sRI should block ∼160,000 pg/ml TNF-α, which is 10-fold more TNF-α than that released by MNLs as they migrate through HUVECs. To quantify the TNF-α released by migrating MNLs, aliquots of the supernatant were collected at the end of the 2-h migration assay and tested with a chemiluminescent TNF-α kit (R&D Systems).

To purify lymphocytes, monocytes were allowed to phagocytose iron beads and were removed with a magnet. Residual monocytes were removed by adherence to plastic for 30 min at 37°C. This protocol leaves <0.1% CD14+ monocytes, as determined by flow cytometry (18, 31). To prepare a T cell-depleted population, we incubated MNLs with anti-CD3-coated iron beads (Dynal Biotech, Brown Deer, WI), at a ratio of 20 beads/T cell, for 30 min on ice. Bead-coated T cells were removed with a magnet, and the cycle was repeated. The resulting population contained <1% CD3-positive T cells by flow cytometry.

Normal donor MNLs were infected with HIV-1 at a multiplicity of infection of 0.01 in the absence of mitogens or added IL-2 (19). This protocol generates ∼0.1–0.01% provirus-positive cells, a fraction comparable to what we found in HIV+ patients’ freshly isolated MNLs (19). Two strains of HIV-1 were used, the M-tropic UC5 and the T-tropic Phlp (a field isolate propagated in MT2 cells). Cells inoculated in this manner do not produce virus until further stimulated, and there is no detectable p24 in their cytoplasm as measured by flow cytometry. After 3 days, the HIV-inoculated MNLs were washed, resuspended in fresh medium, and used in transmigration studies. Cell viability at that time was >95%, determined with trypan blue. Just before the migration assay, the 3-day-old HIV-inoculated MNLs were mixed with an equal number of freshly prepared MNLs from the same donor as a source of fresh resting monocytes. This is necessary because monocytes cultured for 3 days begin to differentiate toward macrophages and are no longer representative of circulating monocytes from peripheral blood. We estimated the frequency of infectious cells in the starting and migratory populations with the limiting dilution assay in which serial dilutions of MNLs are added to duplicate wells containing 0.5 × 106 PHA-stimulated allogeneic lymphoblasts in the presence of IL-2 (17). After 7 days we measured the p24 Ag released into the supernatant (ELISA; Coulter Hialeah, FL) to determine which cultures had received MNLs with infectious virus. The frequency of infected cells was calculated by reference to published tables (17).

MNLs were added to medium flowing at 2 dynes/cm2 across a HUVEC monolayer in a parallel plate flow chamber (34, 35). HUVEC monolayers used to assay migration under flow conditions were routinely prestimulated overnight with 10 ng/ml TNF-α, because leukocytes will not migrate across unstimulated endothelium under these flow conditions. The endothelial monolayers were washed before the assay began and were videotaped for 10 min after the leukocytes began to flow over the surface. Leukocytes that rolled, adhered, or migrated were identified as previously described (34, 35), except that we used more stringent criteria to define adherent cells (full arrest for 1 min rather than 5 s). We did not observe the formation of strings of cells as has been reported previously (36, 37), but did note that several cells would often migrate through the same spot on the endothelial monolayer. Migrating cells typically assumed a polarized shape, extended a cellular process, funneled through a narrow constriction, and then assumed a phase-dark appearance as they spread below the monolayer. After migration, they localized in a focal plane below the endothelial surface and remained phase-dark as they moved below the endothelial cells.

To verify that MNLs were migrating below the endothelium, we stained MNLs with an FITC-based intracytoplasmic dye (CellTracker Green; Molecular Probes, Eugene, OR) before the migration. After migration, we fixed the endothelial monolayers that provided the base of the flow chamber with paraformaldehyde, then added PE-conjugated anti-CD45 to the apical surface for 5 min. Cells adherent to the apical side of the HUVECs gave both an FITC and a PE signal. Cells that had migrated through the endothelium gave off an FITC fluorescence color resulting from the cytoplasmic staining with CellTracker Green before migration. However, because they were shielded from PE-anti-CD45 Ab by the overlying endothelium, they did not show any PE signal. The subendothelial position of the FITC-positive, PE-negative population was further confirmed by demonstrating that they were located in the focal plane beneath that occupied by the FITC-positive, PE-positive cells that adhered to the apical side of the HUVECs.

Flow cytometric analyses were performed as previously described (38) on an EPICS XL cytometer calibrated daily with Flow-Check fluorospheres (Beckman Coulter, Miami, FL). Photomultiplier tube voltages were adjusted each day to a target range of fluorescent intensities by means of Flow-Set fluorospheres (Beckman Coulter) to remove variability attributable to day-to-day fluctuations in instrument performance. mAbs (Beckman Coulter) included anti-CD14 (clone M5E2), anti-CD11b (CR3, clone Bear1), CD49d (clone HP2/1), CD49e (clone Sam1), CD62L (clone DREG 56), CD45 (clone J.33), CD4 (clone SFCI12T4D11), and CD8 (clone B9.11). Isotype control Abs were obtained from BD Biosciences (San Jose, CA) as were mAbs to HLA-DR (clone L243), CD3 (clone UCHT1), CD19 (clone J4.119), CD56 (NCAM16.2), and CD11a (clone G-25.2). We also used anti-CD14 from DakoCytomation (clone TUK4; Carpinteria, CA).

Statistical analysis and graphic representations of data were performed using STATISTICA 6 (StatSoft, Tulsa, OK).

In the presence of FNf, MNLs adhere in greater numbers to endothelial cells, leaving significantly fewer nonadherent cells in the overlying medium. Monocyte adherence, reflected as a depletion of nonadherent cells, was significantly increased within 20 min, and lymphocyte adherence was significantly increased within 40 min after FNf treatment (p = 0.01, by Mann-Whitney U test; Fig. 1). When allowed to remain with the endothelium for 2 h, stimulation with as little as 1 μg/ml FNf significantly increased the numbers of lymphocytes and monocytes that subsequently migrated through the HUVEC monolayer (p < 0.05, by Mann-Whitney U; Fig. 2). Doses of FNf >20 μg/ml caused a small, but statistically significant, additional increase in the numbers of migratory monocytes, whereas lymphocyte migration reached a plateau with 20 μg/ml FNf. Migration was not affected by intact FN (in doses up to 80 μg/ml) or by the CS-1 fragment of FN (in doses up to 250 μg/ml; Fig. 3). The CS-1 fragment binds VLA-4, but does not include the RGD sequence and does not bind VLA-5 (33). FNf increased the migration of both CD4 and CD8 T cells (Fig. 4; p = 0.05, by Mann-Whitney U test). On the average, two-thirds of the migratory T cells express CD45RO, and one-third express CD45RA, and the migration of both increases with FNf stimulation (not shown).

FIGURE 1.

FNf induced adherence of MNLs to HUVECs. MNLs were added to the medium above confluent HUVEC monolayers in the presence (▪) or the absence (○) of 5 μg/ml FNf. Visual inspection confirmed that all MNLs had settled onto the surface of the HUVECs by 15 min. At 20, 40, and 60 min, MNLs that had not adhered to HUVECs were recovered with gentle washing and enumerated by flow cytometry. Error bars are the SD of triplicate determinations. FNf treatment induced a significant decrease in the number of nonadherent leukocytes, indicating a significant increase in the number adherent to HUVECs within 20 min for monocytes and within 40 min for lymphocytes (∗, p = 0.05, by Mann-Whitney U test).

FIGURE 1.

FNf induced adherence of MNLs to HUVECs. MNLs were added to the medium above confluent HUVEC monolayers in the presence (▪) or the absence (○) of 5 μg/ml FNf. Visual inspection confirmed that all MNLs had settled onto the surface of the HUVECs by 15 min. At 20, 40, and 60 min, MNLs that had not adhered to HUVECs were recovered with gentle washing and enumerated by flow cytometry. Error bars are the SD of triplicate determinations. FNf treatment induced a significant decrease in the number of nonadherent leukocytes, indicating a significant increase in the number adherent to HUVECs within 20 min for monocytes and within 40 min for lymphocytes (∗, p = 0.05, by Mann-Whitney U test).

Close modal
FIGURE 2.

FNf induced transendothelial migration of monocytes and lymphocytes. MNLs were added to confluent monolayers of HUVECs in the presence of increasing quantities of FNf for 2 h, and the migratory lymphocytes and monocytes were enumerated by flow cytometry. Data from three experiments are included, each normalized to the number of MNLs in that experiment that migrated in the absence of FNf. FNf significantly increased lymphocyte and monocyte migration at all doses tested (p < 0.05, by Mann-Whitney U test).

FIGURE 2.

FNf induced transendothelial migration of monocytes and lymphocytes. MNLs were added to confluent monolayers of HUVECs in the presence of increasing quantities of FNf for 2 h, and the migratory lymphocytes and monocytes were enumerated by flow cytometry. Data from three experiments are included, each normalized to the number of MNLs in that experiment that migrated in the absence of FNf. FNf significantly increased lymphocyte and monocyte migration at all doses tested (p < 0.05, by Mann-Whitney U test).

Close modal
FIGURE 3.

Effect of intact FN and different FN fragments on MNL migration. MNLs were allowed to migrate across confluent monolayers of HUVECs in the presence of medium alone (no Rx), 5 μg/ml FNf (110-kDa fragment), 20 μg/ml intact FN, or 250 μg/ml CS-1 fragment of FN. After 2 h, the numbers of migratory monocytes (▦) and lymphocytes (▪) were assessed by flow cytometry. Error bars are the SD of triplicate determinations. FNf caused a significant increase in leukocyte migration (∗, p = 0.05, by Mann-Whitney U test), but intact FN and the CS-1 fragment did not.

FIGURE 3.

Effect of intact FN and different FN fragments on MNL migration. MNLs were allowed to migrate across confluent monolayers of HUVECs in the presence of medium alone (no Rx), 5 μg/ml FNf (110-kDa fragment), 20 μg/ml intact FN, or 250 μg/ml CS-1 fragment of FN. After 2 h, the numbers of migratory monocytes (▦) and lymphocytes (▪) were assessed by flow cytometry. Error bars are the SD of triplicate determinations. FNf caused a significant increase in leukocyte migration (∗, p = 0.05, by Mann-Whitney U test), but intact FN and the CS-1 fragment did not.

Close modal
FIGURE 4.

Effect of FNf on the migration of CD4 and CD8 T cells. MNLs were allowed to migrate across confluent HUVEC monolayers in the presence or the absence of FNf (5 μg/ml). After 2 h, the migratory MNLs were harvested, stained with Abs to CD3 and CD4 or CD8, and enumerated by flow cytometry. The results of two experiments under each condition are shown. Error bars are the SD of triplicate determinations. FNf treatment induced a significantly greater number of both CD4 and CD8 T cells to migrate in each case (p = 0.05, by Mann-Whitney U test).

FIGURE 4.

Effect of FNf on the migration of CD4 and CD8 T cells. MNLs were allowed to migrate across confluent HUVEC monolayers in the presence or the absence of FNf (5 μg/ml). After 2 h, the migratory MNLs were harvested, stained with Abs to CD3 and CD4 or CD8, and enumerated by flow cytometry. The results of two experiments under each condition are shown. Error bars are the SD of triplicate determinations. FNf treatment induced a significantly greater number of both CD4 and CD8 T cells to migrate in each case (p = 0.05, by Mann-Whitney U test).

Close modal

Exposure to FNf also increased, in a dose-dependent manner, the quantity of TNF-α released by MNLs as they migrate (Fig. 5,A). HUVEC monolayers exposed to FNf without addition of MNLs produced <1 ng/ml TNF-α during this 2-h period (data not shown). FNf stimulation of monocyte-depleted vs T cell-depleted populations verified that FNf induced TNF-α release from monocytes, but not T cells (Fig. 5 B). Intact FN and the CS-1 fragment did not induce TNF-α release (data not shown).

FIGURE 5.

Effect of FNf on TNF-α release by MNLs. A, MNLs were added to confluent monolayers of HUVECs in the presence of increasing quantities of FNf. At the end of 2 h, we measured the quantity of TNF-α released into the supernatant above the HUVECs. Error bars indicate ±SD of triplicate determinations. FNf significantly increased the release of TNF-α at all doses tested (p < 0.05, by Mann-Whitney U test). The results shown are representative of those obtained with cells from three normal donors. B, Unfractionated MNLs, monocyte-depleted MNLs, and T cell-depleted MNLs were stimulated with FNf (5 μg/ml) in Teflon jars, and the quantity of TNF-α released in the supernatant was assayed after 24 h. FNf induced release of TNF-α from monocytes, but not T cells. An additional control, consisting of reconstituted MNLs using monocyte-depleted MNLs (e.g., T cells) and T cell-depleted MNLs (e.g., monocytes) is included to show the effects of cell handling. It illustrates that the lower amounts of TNF-α released by purified monocytes compared with unfractionated cells is due to the process of cell handling, because readdition of T cells (in the form of monocyte-depleted MNLs) does not increase the quantity of TNF-α released. In the absence of FNf stimulation, monocytes released <60 pg/ml TNF-α (not shown).

FIGURE 5.

Effect of FNf on TNF-α release by MNLs. A, MNLs were added to confluent monolayers of HUVECs in the presence of increasing quantities of FNf. At the end of 2 h, we measured the quantity of TNF-α released into the supernatant above the HUVECs. Error bars indicate ±SD of triplicate determinations. FNf significantly increased the release of TNF-α at all doses tested (p < 0.05, by Mann-Whitney U test). The results shown are representative of those obtained with cells from three normal donors. B, Unfractionated MNLs, monocyte-depleted MNLs, and T cell-depleted MNLs were stimulated with FNf (5 μg/ml) in Teflon jars, and the quantity of TNF-α released in the supernatant was assayed after 24 h. FNf induced release of TNF-α from monocytes, but not T cells. An additional control, consisting of reconstituted MNLs using monocyte-depleted MNLs (e.g., T cells) and T cell-depleted MNLs (e.g., monocytes) is included to show the effects of cell handling. It illustrates that the lower amounts of TNF-α released by purified monocytes compared with unfractionated cells is due to the process of cell handling, because readdition of T cells (in the form of monocyte-depleted MNLs) does not increase the quantity of TNF-α released. In the absence of FNf stimulation, monocytes released <60 pg/ml TNF-α (not shown).

Close modal

To determine whether TNF-α was responsible for the FNf-driven increase in MNL migration, we added TNF sRI, a highly avid soluble receptor for TNF-α that abolishes the biological activity of this cytokine (18, 31). TNF sRI completely blocked the migration-enhancing effects of FNf (Fig. 6). Addition of nonspecific IgG as a control fusion protein for the TNF sRI/Fc chimera had no effect on lymphocyte or monocyte migration (not shown).

FIGURE 6.

Role of TNF-α in FNf-induced migration. MNLs were allowed to migrate across confluent monolayers of HUVECs in the presence of FNf with and without TNF sRI to block TNF-α effects. The number of lymphocytes (▪) and monocytes (▦) that migrated at the end of 2 h are shown. Error bars show ±1 SD of triplicate determinations. The significant increase in migration of lymphocytes and monocytes caused by FNf was blocked by TNF sRI (0.64 μg/ml; p < 0.05, by Mann-Whitney U test). The results shown are representative of those obtained with MNLs from two normal donors.

FIGURE 6.

Role of TNF-α in FNf-induced migration. MNLs were allowed to migrate across confluent monolayers of HUVECs in the presence of FNf with and without TNF sRI to block TNF-α effects. The number of lymphocytes (▪) and monocytes (▦) that migrated at the end of 2 h are shown. Error bars show ±1 SD of triplicate determinations. The significant increase in migration of lymphocytes and monocytes caused by FNf was blocked by TNF sRI (0.64 μg/ml; p < 0.05, by Mann-Whitney U test). The results shown are representative of those obtained with MNLs from two normal donors.

Close modal

TNF-α could increase migration by activating MNLs, activating endothelium, or both; our data indicate that it activates both. FNf and TNF-α modulate the expression of leukocyte adhesion molecules known to be involved in transendothelial migration and translocation through protein matrices. Stimulation of monocytes with either TNF-α or FNf for 1 h increased the quantity of CD11b on their membranes while decreasing the expression of CD62L (Fig. 7). We have previously shown that FNf also induces monocytes to shed CD49e (23). Lymphocytes exposed to TNF-α or FNf expressed diminished quantities of CD11a and CD49e (Fig. 7). The expression of CD49d did not change significantly on the MNLs exposed to FNf or TNF-α.

FIGURE 7.

Effect of FNf or TNF-α on leukocyte adhesion molecules. MNLs were treated with medium (▦), FNf (5 μg/ml; ▨), or TNF (10 ng/ml; □) for 1 h at 37°C and then stained for adhesion markers: CD11b, CD11a, CD62L, and CD49e. Cells were treated in a Teflon jar, processed in siliconized tubes, and washed with gelatin-containing buffers throughout to prevent cell loss through adhesion. Data shown are the density of the marker (mcf, mean channel fluorescence) or percentage of CD14+ monocytes or CD45+ CD14 lymphocytes that was positive for the marker. Delimiters were set using MNLs stained with nonspecific isotype controls. Results are representative of three experiments.

FIGURE 7.

Effect of FNf or TNF-α on leukocyte adhesion molecules. MNLs were treated with medium (▦), FNf (5 μg/ml; ▨), or TNF (10 ng/ml; □) for 1 h at 37°C and then stained for adhesion markers: CD11b, CD11a, CD62L, and CD49e. Cells were treated in a Teflon jar, processed in siliconized tubes, and washed with gelatin-containing buffers throughout to prevent cell loss through adhesion. Data shown are the density of the marker (mcf, mean channel fluorescence) or percentage of CD14+ monocytes or CD45+ CD14 lymphocytes that was positive for the marker. Delimiters were set using MNLs stained with nonspecific isotype controls. Results are representative of three experiments.

Close modal

To evaluate the effect of TNF-α on lymphocytes vs endothelial cells in terms of migration, we treated them separately or together before or during the migration assay (Fig. 8). When assaying the effect of TNF-α on lymphocyte migration, monocytes were removed because they are a constitutive source of TNF-α (19, 31). Pulse treatment of purified lymphocytes with TNF-α before the migration assay induced a 1.86 ± 0.6-fold increase in migration (mean ± SD of nine donors; p = 0.015, by t test), indicating that TNF-α can directly affect lymphocyte migration (Fig. 8, compare result B to A). Pulse treatment was more effective than continuous treatment of the lymphocytes during the migration (Fig. 8, result B vs C; p = 0.05, three donors, by Mann-Whitney U test). Activation of the endothelium alone with TNF-α also increased migration (Fig. 8, compare result D to A). However, even when the endothelium was maximally stimulated by prior exposure to TNF-α for 18 h, addition of more TNF-α during the migration assay caused a small, but significant, increase in the number of lymphocytes that passed through the endothelium (Fig. 8, compare result E or F to D).

FIGURE 8.

TNF-α stimulates lymphocyte migration directly. Purified lymphocytes were treated with TNF-α for 10 min (Pretreat) or throughout the migration assay (During migration). Lymphocytes were added to unstimulated HUVECs (−) or HUVECs that had been pretreated with TNF-α overnight and washed (Pretreat). Bars indicate the percentage of added lymphocytes that migrated (mean ± SD of triplicate determinations). The experiment shown is representative of three experiments with stimulated endothelium and nine experiments with unstimulated endothelium. ∗, Significantly different from untreated cells using pooled normalized data from all experiments (p < 0.05, by t test).

FIGURE 8.

TNF-α stimulates lymphocyte migration directly. Purified lymphocytes were treated with TNF-α for 10 min (Pretreat) or throughout the migration assay (During migration). Lymphocytes were added to unstimulated HUVECs (−) or HUVECs that had been pretreated with TNF-α overnight and washed (Pretreat). Bars indicate the percentage of added lymphocytes that migrated (mean ± SD of triplicate determinations). The experiment shown is representative of three experiments with stimulated endothelium and nine experiments with unstimulated endothelium. ∗, Significantly different from untreated cells using pooled normalized data from all experiments (p < 0.05, by t test).

Close modal

To evaluate whether TNF-α also influences monocyte migration, we pulse-treated unfractionated MNLs. TNF-α treatment increased monocyte migration by 2.08 ± 1.02-fold (p = 0.008) and migration of the accompanying lymphocytes by 1.95 ± 0.68-fold (mean ± SD of nine donors; p = 0.008, by paired t test compared with untreated MNLs; Fig. 9). Dose-response and kinetic analyses performed to identify the optimum quantity and timing of treatment showed that TNF-α caused a significant increase in MNL migration when used at doses as low at 1 pg/ml and that exposure of the leukocytes to TNF-α for <1 min was sufficient to stimulate migration (data not shown).

FIGURE 9.

TNF-α stimulates lymphocyte and monocyte migration. MNLs or lymphocytes were treated with TNF-α, washed, and allowed to migrate for 2 h across unstimulated HUVECs. Bars indicate the percentage of added cells that migrated (mean ± SD triplicate wells). TNF-α treatment of unfractionated MNLs caused a significant increase in monocyte and lymphocyte migration (p < 0.05, by Mann-Whitney U test). The results shown are representative of those from nine donors.

FIGURE 9.

TNF-α stimulates lymphocyte and monocyte migration. MNLs or lymphocytes were treated with TNF-α, washed, and allowed to migrate for 2 h across unstimulated HUVECs. Bars indicate the percentage of added cells that migrated (mean ± SD triplicate wells). TNF-α treatment of unfractionated MNLs caused a significant increase in monocyte and lymphocyte migration (p < 0.05, by Mann-Whitney U test). The results shown are representative of those from nine donors.

Close modal

Transendothelial migration in vivo involves multiple steps: rolling, arrest, and spreading over the endothelial cells, and finally translocation through the vascular barrier (39). To identify which of these steps was affected by TNF-α, we monitored the behavior of TNF-α-treated cells as they interacted with HUVECs under flow conditions where each of these steps can be measured individually. Over a 10-min period, we counted the cells that rolled but never adhered, the cells that adhered without migrating, and the cells that migrated (many of which rolled and/or adhered before they migrated). TNF-α significantly increased the number of cells that migrated, but did not change the number that only rolled or adhered without migrating (Fig. 10). Fig. 10 shows results with pulse-treated cells. The same outcome occurred when TNF-α was present throughout the assay. The continuous presence of TNF-α caused a 1.68 ± 0.64-fold increase in the numbers of migratory cells (mean ± SD; seven donors; p = 0.0009, by paired t test) without affecting the number of cells that only rolled or adhered to the HUVECs.

FIGURE 10.

TNF-α stimulates lymphocyte migration under flow conditions. Purified lymphocytes were treated with 10 ng/ml TNF-α (▪) or medium alone (□) for 10 min, washed, and allowed to interact with HUVECs under shear stress conditions. Bars indicate the numbers of cells that rolled, adhered, or migrated within 10 min (mean ± SD). TNF-α induced a significant increase in the number of cells that migrated, but did not change the number that only rolled or adhered. Results are representative of those obtained with two donors’ MNLs pretreated with TNF-α before the migration and seven donors’ MNLs exposed to TNF-α throughout the migration.

FIGURE 10.

TNF-α stimulates lymphocyte migration under flow conditions. Purified lymphocytes were treated with 10 ng/ml TNF-α (▪) or medium alone (□) for 10 min, washed, and allowed to interact with HUVECs under shear stress conditions. Bars indicate the numbers of cells that rolled, adhered, or migrated within 10 min (mean ± SD). TNF-α induced a significant increase in the number of cells that migrated, but did not change the number that only rolled or adhered. Results are representative of those obtained with two donors’ MNLs pretreated with TNF-α before the migration and seven donors’ MNLs exposed to TNF-α throughout the migration.

Close modal

Considering that FNf modulates the migration of monocytes in FN-rich tissue matrices (23), we investigated the effect of FNf treatment on the behavior of cells after they had passed through the endothelium. Untreated MNLs tended to migrate deep into the collagen pad during the 2-h period, whereas FNf caused MNLs to accumulate just below the endothelial monolayer (Fig. 11). After FNf treatment, there were significantly more leukocytes in the first 100 μm of the collagen pad and significantly fewer in the region 300–400 μm below the endothelial monolayer (p < 0.05, by t test).

FIGURE 11.

Effect of FNf on depth of migration below endothelium. MNLs were allowed to migrate across confluent monolayers of HUVECs and into the collagen pad in the presence (dotted line) or the absence (solid line) of FNf. At the end of 2 h, the number of migratory cells that accumulated at consecutive 100-μm intervals below the endothelium were counted. Results are presented as the fraction of the total migratory population at each level. Error bars indicate ±SD of nine replicate determinations. ∗, Significantly different values (p < 0.05, by t test).

FIGURE 11.

Effect of FNf on depth of migration below endothelium. MNLs were allowed to migrate across confluent monolayers of HUVECs and into the collagen pad in the presence (dotted line) or the absence (solid line) of FNf. At the end of 2 h, the number of migratory cells that accumulated at consecutive 100-μm intervals below the endothelium were counted. Results are presented as the fraction of the total migratory population at each level. Error bars indicate ±SD of nine replicate determinations. ∗, Significantly different values (p < 0.05, by t test).

Close modal

Monocytes that encounter proinflammatory stimuli after they migrate through vascular barriers in vitro may differentiate into dendritic cells and reverse-migrate through the endothelial monolayer to the apical surface (17, 40). This reverse migration is thought to model the movement of dendritic cells from tissue into lymphatics en route to lymph nodes (40). To evaluate whether FNf treatment enhances reverse migration, we washed away the nonmigratory cells and refed the cultures. Two days later we collected the cells that emerged from the apical side of the HUVEC monolayer. We found that FNf increased both the number of HLA-DRdim monocytoid cells that reverse-migrated and the fraction that expressed high levels of HLA-DR, a characteristic of dendritic cells (Fig. 12). The number of lymphocytic cells (CD3-positive, CD19-positive, or CD56-positive cells) that reverse-migrated was not significantly increased by FNf treatment.

FIGURE 12.

Effect of FNf on differentiation of reverse-migrating dendritic cells. MNLs were allowed to migrate across confluent HUVEC monolayers in the presence or the absence of FNf. Nonmigratory cells were washed away after 2 h, cultures were refed, and 2 days later, MNLs reverse-migrating to the apical side of the HUVECS were collected. Results show the numbers of lymphocytic (CD3+, CD19+, or CD56+), monocytoid (CD3CD19CD56HLA-DRdim), and dendritic-type cells (CD3CD19CD56HLA-DRbright). Error bars indicate the results of triplicate determinations, and results are representative of three donors. The number of reverse-migrating dendritic cells was significantly increased in the presence of FNf (p = 0.05, by Mann-Whitney U test).

FIGURE 12.

Effect of FNf on differentiation of reverse-migrating dendritic cells. MNLs were allowed to migrate across confluent HUVEC monolayers in the presence or the absence of FNf. Nonmigratory cells were washed away after 2 h, cultures were refed, and 2 days later, MNLs reverse-migrating to the apical side of the HUVECS were collected. Results show the numbers of lymphocytic (CD3+, CD19+, or CD56+), monocytoid (CD3CD19CD56HLA-DRdim), and dendritic-type cells (CD3CD19CD56HLA-DRbright). Error bars indicate the results of triplicate determinations, and results are representative of three donors. The number of reverse-migrating dendritic cells was significantly increased in the presence of FNf (p = 0.05, by Mann-Whitney U test).

Close modal

HIV-1-infected MNLs display increased quantities of CD11a, CD18, CD29, and CD49d and become more adherent to endothelial cells, more likely to engage in homotypic adhesive interactions, and more likely to migrate across endothelial barriers (19, 20). Activated normal donor lymphocytes are also more responsive to the migration-enhancing effects of TNF-α (31). Consequently, we postulated that stimulation with FNf might increase migration of HIV-infected cells to a greater degree than uninfected cells. To test this hypothesis, we infected normal donor MNLs with the T-tropic (Phlp) or the M-tropic (UC5) strain of HIV-1 for 2 days without exogenous mitogens or cytokines. Under these conditions, ∼0.01–0.1% of the leukocytes became infected. These HIV-infected MNLs were then allowed to migrate across endothelial monolayers in the presence or the absence of FNf. We measured the numbers of HIV-infected MNLs recovered from the collagen pad by means of a limiting dilution assay in which MNLs were incubated with highly susceptible PHA-stimulated lymphoblasts in the presence of IL-2. The limiting dilution assay is designed to detect any cell capable of transmitting HIV, including leukocytes with latent virus that can be reactivated with stimulation. With both T-tropic and M-tropic HIV-1 strains, FNf treatment increased the total number of migratory MNLs and the number of HIV-infected MNLs recovered from the collagen pads 2–24 h later (Fig. 13). Fig. 13 shows the effects of FNf on these infected MNLs, expressed as the fold increase in the number of migratory cells after FNf compared with the increase in the number of infected MNLs treated with medium alone. FNf treatment induced a 2- to 5-fold increase in the number of HIV-infected cells that migrated (p = 0.02 for cells collected after 24 h; p = 0.008 for 2 and 24 h harvests analyzed together; by Wilcoxon test). At the same time, FNf induced only a <2-fold increase in the total number of cells that migrated. If infected and uninfected cells responded in the same way to FNf, then the fold increase in migration caused by FNf should be the same for HIV-infected and uninfected cells. However, as shown in Fig. 13, the increase in recovery of HIV-infected cells among the migratory cells consistently exceeded the increase in the total number of MNLs recovered from the subendothelial compartment. This was equally true for cultures that were harvested within 2 h of migration (see inset, Fig. 13). This demonstrates that FNf preferentially promotes the migration of HIV-1-infected cells. The original cell populations in these experiments contained varying numbers of infected cells, ranging from 82–1122/million MNLs (Table I). However, the level of infection did not appear to influence the total number of MNLs, the number of HIV-infected cells that migrated, or the response of these cells to stimulation with FNf.

FIGURE 13.

Effect of FNf on migration of HIV-infected MNLs. MNLs from six normal donors were infected with HIV (UC5 for no. 1–5 and Phlp for no. 6) for 48 h, then allowed to migrate across confluent monolayers of HUVECs in the presence or the absence of FNf (5 μg/ml for donors 1–3 and 20 μg/ml for donors 4–6). The migratory population was harvested at 2 h (inset for donors 1–3) and 24 h (donors 1–6) to quantify both the total numbers of MNLs (▪) and the numbers carrying infectious HIV (▨). Data are presented as the fold difference in results with FNf-treated cells compared with infected cells treated with medium alone. Actual cell numbers are presented in Table I. FNf induced a significant increase in the migration of HIV-infected MNLs (p = 0.008 for all time points; p = 0.02 for 24 h harvests alone; by Wilcoxon test).

FIGURE 13.

Effect of FNf on migration of HIV-infected MNLs. MNLs from six normal donors were infected with HIV (UC5 for no. 1–5 and Phlp for no. 6) for 48 h, then allowed to migrate across confluent monolayers of HUVECs in the presence or the absence of FNf (5 μg/ml for donors 1–3 and 20 μg/ml for donors 4–6). The migratory population was harvested at 2 h (inset for donors 1–3) and 24 h (donors 1–6) to quantify both the total numbers of MNLs (▪) and the numbers carrying infectious HIV (▨). Data are presented as the fold difference in results with FNf-treated cells compared with infected cells treated with medium alone. Actual cell numbers are presented in Table I. FNf induced a significant increase in the migration of HIV-infected MNLs (p = 0.008 for all time points; p = 0.02 for 24 h harvests alone; by Wilcoxon test).

Close modal
Table I.

Effect of FNf on migration total MNLs and HIV-infected MNLsa

Donor No.Migration Time (h)No. of HIV-Infected CellsbTotal MNLsc Migrating after Sham Treatment
Migrating after Sham TreatmentMigrating after FNf TreatmentPer 106 Starting MNLs
95 525 420 1,054,430 
24 1476 2350 420 3,188,450 
107 276 82 1,195,810 
24 454 872 82 2,005,840 
21 123 1,122 2,308,320 
24 45 118 1,122 2,518,700 
24 46 105 420 1,626,160 
24 18 206 2,837,600 
24 18 1,122 6,058,400 
Donor No.Migration Time (h)No. of HIV-Infected CellsbTotal MNLsc Migrating after Sham Treatment
Migrating after Sham TreatmentMigrating after FNf TreatmentPer 106 Starting MNLs
95 525 420 1,054,430 
24 1476 2350 420 3,188,450 
107 276 82 1,195,810 
24 454 872 82 2,005,840 
21 123 1,122 2,308,320 
24 45 118 1,122 2,518,700 
24 46 105 420 1,626,160 
24 18 206 2,837,600 
24 18 1,122 6,058,400 
a

These same data are shown in Fig. 13, but presented there as the fold increase in FNf-treated cultures compared to sham-treated cultures.

b

To calculate the numbers of HIV-infected cells that migrated under each condition, we multiplied the total number of migratory cells by the frequency of HIV-infected MNLs in that population, as measured with the limiting dilution assay. For the starting population, data are reported as the number of HIV-infected cells per million MNLs.

c

In each experiment, 11 HUVEC wells, representing 11 million starting MNLs, were treated with FNf, and an equal number of wells were treated with medium. Migratory cells from the 11 replicate wells were pooled for the limiting dilution assay. The data shown here are the numbers of sham-treated cells that migrated; Fig. 13 shows the fold increase induced by FNf treatment.

These studies were undertaken to investigate why some, but by no means all, HIV-1-infected men whose blood contains productively infected leukocytes have circulating MNLs that can transport infectious virus across endothelial barriers in vitro and disseminate HIV-1 among a large number of uninfected lymphocytes and monocytes drawn to the same sites, presumably by agents that caused the infected cells to migrate in the first place (17). The clinical significance of these observations was initially obscure until, with continued observation of this cohort of patients, it became apparent that complications associated with progression of HIV-1 infection occurred more often in those subjects whose MNLs exhibited this behavior (17). Recognition that FN fragments are present, at least intermittently, in the plasma of many HIV-1-infected patients (38) and that these FNf are among the triggers that cause monocytes to release cytoplasmic stores of TNF-α (26, 27, 28) caused us to consider the possibility that FNf may be one of the agents that induces MNLs, particularly HIV-1-infected MNLs, to migrate across endothelial barriers.

The experiments presented in this report suggest that an important effect of FNf is to stimulate MNLs to release TNF-α. The central role of TNF-α in driving MNL migration is confirmed by demonstrating that the increase in MNL migration induced by FNf is blocked by TNF sRI. A focal increase in TNF-α has many effects. It up-regulates endothelial expression of selectins, ICAMs, and VCAM-1, facilitating adhesive interactions that promote leukocyte transendothelial migration (31). The experiments presented in this report suggest that the TNF-α may also directly stimulate MNL migration, even when the endothelium is not activated and even under shear stress conditions. One mechanism may be the modulation of adhesion molecules involved in transendothelial migration and translocation across protein matrices. TNF-α and FNf induce monocytes to shed CD49e and CD62L and to increase the expression of CD11b. They also induce lymphocytes to decrease their expression of CD11a and CD49e, but have little effect on CD49d. One can visualize several ways in which changes in adhesion molecule expression might affect leukocyte motility. Migration involves a dynamic process of attachment and detachment. It is possible that a diminution in cell surface adhesins may prevent a cell from being tethered and allow it to migrate more easily. The phenotypic data shown in Fig. 7 are a single snapshot after 60 min of stimulation. These markers continue to change with time after exposure to FNf. For example, at 24 h, monocyte CD62L and CD11b return to normal, but CD49d and CD49e become elevated on FNf-treated cells. Lymphocyte CD11a and CD49e return to normal, but CD11b decreases to levels below those seen on untreated MNLs (not shown).

Stimulation with FNf increased the recovery of HIV-infected cells among the migratory population. Infection with HIV activates leukocytes. HIV-infected cells express increased quantities of integrins (20, 41, 42, 43), adhere in greater numbers to endothelial monolayers (20, 41, 42, 43), and have a competitive advantage over uninfected cells when migrating across monolayers of resting endothelium (19). It is not surprising then that HIV-infected cells would also migrate preferentially over endothelium that has been activated by TNF-α released in response to FNf. A similar effect was seen using endothelium stimulated with LPS to up-regulate vascular adhesion molecules (data not shown). In one experiment, 12-fold more HIV-infected MNLs traversed the LPS-activated endothelium compared with resting endothelium, whereas the total number cells that migrated was only 13% higher for LPS-stimulated HUVECs (325 HIV-infected cells and 1.58 million total MNLs migrated across unstimulated HUVECs compared with 4491 HIV-infected cells and 1.79 million total MNLs migrating across LPS-stimulated HUVECs).

In addition to enhancing the transendothelial migration of HIV-infected cells, FNf may have stabilized cell surface HIV, thereby enhancing transmission of virus to the PHA-stimulated lymphoblasts. Recent reports suggest that diverse forms of matrix FN can stabilize virions of HIV-1 on the surface of infected cells and help facilitate the transmission of infectious virus (24, 25). We have also found that addition of FNf to cultures of in vitro infected MNLs stimulated with IL-2 increases by 2- to 4-fold the quantity of p24 generated over the next 3 days (data not shown).

Leukocyte integrins are involved in interactions with FN. Leukocytes can interact with immobilized FN using both VLA-4 (CD29/CD49d) and VLA-5 (CD29/CD49e). The 110-kDa FNf fragment used in these studies includes the RGD sequence and binds to CD49e (27). The other FN fragment we used, CS-1, does not contain the RGD sequence and does not bind to VLA-5. The CS-1 fragment is closer to the C terminus of the FN molecule and binds to VLA-4. Neither the CS-1 fragment nor intact FN induced TNF-α release or affected the migration of MNLs.

After exposure to FNf, monocytes shed CD49e and are less able to migrate through collagenous matrices rich in native FN (23). FNf-induced loss of CD49e may also explain the failure of treated MNLs to penetrate deeply into FN-collagen matrices below the endothelial barrier in the present report. Colocalization of large numbers of infected and uninfected cells in a compact site is likely to promote interactions that facilitate the transfer of HIV-1. Accumulation of these foci of infected and uninfected MNLs just under the endothelial surface may also enhance reverse migration of the subset of infected MNLs that develop characteristics of dendritic cells. To the extent that this process is replicated in vivo, these experiments suggest that circulating FNf in HIV-1-infected patients could help disseminate HIV-1 in extravascular soft tissue depots within an infected host. Whether the development of these perivascular foci help to protect the infected cells within from the actions of antiretroviral drugs remains to be determined.

Cell surface display of HLA-DR, a phenotypic characteristic of monocytes that have begun to differentiate into dendritic cells (44), is prominently seen on FNf-stimulated monocytes, especially those that reverse-migrated. Release of TNF-α, as a result of FNf stimulation, may induce monocytes to differentiate toward dendritic cells (40, 45). If these developing dendritic cells were to take up HIV-1 from the subendothelial depots, they would be highly effective at infecting T cells that entered the same subendothelial microenvironment (46).

Proteolytic enzymes released by replicating virus and those produced by host inflammatory responses to HIV-1 or to secondary microbial infections may help to break down plasma or tissue FN (41, 47, 48, 49). Prospective studies of the prevalence of circulating FN fragments in HIV-1-infected subjects together with additional studies of their effects on the functioning of cells of the immune response system may deepen our understanding of the ability of these byproducts of the inflammatory response to influence the development of extravascular foci of infection that appear to participate in the cardiac, renal, and CNS complications of HIV-1 infections.

We thank Dr. C. Wayne Smith (Division of Leukocyte Biology, Baylor College of Medicine) for the generous donations of HUVECs.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was supported in part by the Department of Veterans Affairs; National Institutes of Health Grants RO1MH63035, RO1NS32583, RO1AI46285, and AI36211; the core support grant for the Baylor Center for AIDS Research.

3

Abbreviations used in this paper: MNL, mononuclear leukocyte (monocytes and lymphocytes); CS-1, connecting segment; FN, fibronectin; FNf, 110- to 120-kDa FN fragments; TNF sRI, soluble TNFα receptor I.

1
Masliah, E., C. L. Achim, N. Ge, R. DeTeresa, R. D. Terry, C. A. Wiley.
1992
. Spectrum of human immunodeficiency virus-associated neocortical damage.
Ann. Neurol.
32
:
321
.
2
Tyor, W. R., S. L. Wesselingh, J. W. Griffin, J. C. McArthur, D. E. Griffin.
1995
. Unifying hypothesis for the pathogenesis of HIV-associated dementia complex, vacuolar myelopathy, and sensory neuropathy.
J. Acquired Immune Defic. Syndr. Hum. Retrovirol.
9
:
379
. (Rev.).
3
Lewis, W..
2000
. Cardiomyopathy in AIDS: a pathophysiological perspective.
Prog. Cardiovasc. Dis.
43
:
151
.
4
Liu, Q. N., S. Reddy, J. W. Sayre, V. Pop, M. C. Graves, M. Fiala.
2001
. Essential role of HIV type 1-infected and cyclooxygenase 2-activated macrophages and T cells in HIV type 1 myocarditis.
AIDS Res. Hum. Retroviruses
17
:
1423
.
5
Gherardi, P..
1994
. Skeletal muscle involvement in HIV-infected patients.
Neuropathol. Appl. Neurobiol.
20
:
232
.
6
Dalakas, M. C..
1993
. Search for HIV proviral DNA and amplified sequences in the muscle biopsies of patients with HIV polymyositis.
Muscle Nerve
16
:
408
.
7
Bodi, I., A. A. Abraham, P. L. Kimmel.
1994
. Macrophages in human immunodeficiency virus-associated kidney diseases.
Am. J. Kidney Dis.
24
:
762
.
8
Herman, E. S., P. E. Klotman.
2003
. HIV-associated nephropathy: epidemiology, pathogenesis, and treatment.
Semin. Nephrol.
23
:
200
.
9
Kimmel, P. L., L. Barisoni, J. B. Kopp.
2003
. Pathogenesis and treatment of HIV-associated renal diseases: lessons from clinical and animal studies, molecular pathologic correlations, and genetic investigations.
Ann. Intern. Med.
139
:
214
.
10
Moroni, M., A. Repossini, M. Multinelli, A. Lazzarin, C. Antonacci, S. Cristina, C. Negri.
1988
. Focal lymphocytic myocarditis in acquired immunodeficiency syndrome (AIDS): a correlative morphologic and clinical study in 26 consecutive fatal cases.
J. Am. College Cardiol.
12
:
463
.
11
Nuovo, G. J., F. Gallery, P. MacConnell, A. Braun.
1994
. In situ detection of polymerase chain reaction-amplified HIV-1 nucleic acids and tumor necrosis factor-α RNA in the central nervous system.
Am. J. Pathol.
144
:
659
.
12
Giulian, D., K. Vaca, C. A. Noonan.
1990
. Secretion of neurotoxins by mononuclear phagocytes infected with HIV-1.
Science
250
:
1593
.
13
Roux-Lombard, P., C. Modoux, A. Cruchaud, J. M. Dayer.
1989
. Purified blood monocytes from HIV 1-infected patients produce high levels of TNFα and IL-1.
Clin. Immunol. Immunopathol.
50
:
374
.
14
Merrill, J. E., I. S. Chen.
1991
. HIV-1, macrophages, glial cells, and cytokines in AIDS nervous system disease.
FASEB J.
5
:
2391
.
15
Kruman, I. I., A. Nath, M. P. Mattson.
1998
. HIV-1 protein Tat induces apoptosis of hippocampal neurons by a mechanism involving caspase activation, calcium overload, and oxidative stress.
Exp. Neurol.
154
:
276
.
16
Birdsall, H. H., J. Trial, J. A. Hallum, A. L. deJong, L. K. Green, J. C. Bandres, S. A. Smole, A. H. Laughter, R. D. Rossen.
1994
. Phenotypic and functional activation of monocytes in HIV-1 infection: interactions with neural cells.
J. Leukocyte Biol.
56
:
310
.
17
Birdsall, H., E. Siwak, J. Trial, M. Rodriguez-Barradas, A. J. White, S. Weitgrefe, R. Rossen.
2002
. Transendothelial migration of leukocytes carrying infectious HIV-1: an indicator of adverse prognosis.
AIDS
16
:
5
.
18
de Jong, A. L., D. M. Green, J. Trial, H. H. Birdsall.
1996
. Focal effects of mononuclear leukocyte transendothelial migration: TNF-α production by migrating monocytes promotes subsequent migration of lymphocytes.
J. Leukocyte Biol.
60
:
129
.
19
Birdsall, H. H., J. Trial, H. J. Lin, D. M. Green, G. W. Sorrentino, E. B. Siwak, A. L. de Jong, R. D. Rossen.
1997
. Transendothelial migration of lymphocytes from HIV-1 infected donors: a mechanism for extravascular dissemination of HIV-1.
J. Immunol.
158
:
5968
.
20
Rossen, R. D., C. W. Smith, A. H. Laughter, C. A. Noonan, D. C. Anderson, W. M. McShan, M. Y. Hurvitz, F. M. Orson.
1989
. HIV-1-stimulated expression of CD11/CD18 integrins and ICAM-1: a possible mechanism for extravascular dissemination of HIV-1-infected cells.
Trans. Assoc. Am. Physicians
102
:
117
.
21
Weeks, B. S., M. E. Klotman, S. Dhawan, M. Kibbey, J. Rappaport, H. K. Kleinman, K. M. Yamada, P. E. Klotman.
1991
. HIV-1 infection of human T lymphocytes results in enhanced α5β1 integrin expression.
J. Cell Biol.
114
:
847
.
22
Torre, D., G. Ferrario, M. Issi, A. Pugliese, F. Speranza.
1996
. Expression of the α5β1 fibronectin receptor on T lymphocytes of patients with HIV-1 infection.
J. Clin. Pathol.
49
:
733
.
23
Trial, J., R. E. Baughn, J. N. Wygant, B. W. McIntyre, H. H. Birdsall, K. A. Youker, A. Evans, M. L. Entman, R. D. Rossen.
1999
. Fibronectin fragments modulate monocyte VLA-5 expression and monocyte migration.
J. Clin. Invest.
104
:
419
.
24
Greco, G., S. Pal, R. Pasqualini, L. M. Schnapp.
2002
. Matrix fibronectin increases HIV stability and infectivity.
J. Immunol.
168
:
5722
.
25
Tellier, M. C., G. Greco, M. Klotman, A. Mosoian, A. Cara, W. Arap, E. Ruoslahti, R. Pasqualini, L. M. Schnapp.
2000
. Superfibronectin, a multimeric form of fibronectin, increases HIV infection of primary CD4+ T lymphocytes.
J. Immunol.
164
:
3236
.
26
Beezhold, D. H., C. Personius.
1992
. Fibronectin fragments stimulate tumor necrosis factor secretion by human monocytes.
J. Leukocyte Biol.
51
:
59
.
27
Lopez-Armada, M. J., E. Gonzalez, C. Gomez-Guerrero, J. Egido.
1997
. The 80-kD fibronectin fragment increases the production of fibronectin and tumour necrosis factor-α (TNF-α) in cultured mesangial cells.
Clin. Exp. Immunol.
107
:
398
.
28
Neese, L. W., J. E. Standing, E. J. Olson, M. Castro, A. H. Limper.
1994
. Vitronectin, fibronectin, and gp120 antibody enhance macrophage release of TNF-α in response to Pneumocystis carinii.
J. Immunol.
152
:
4549
.
29
Homandberg, G. A., F. Hui, C. Wen, C. Purple, K. Bewsey, H. Koepp, K. Huch, A. Harris.
1997
. Fibronectin-fragment-induced cartilage chondrolysis is associated with release of catabolic cytokines.
Biochem. J.
321
:
751
.
30
Chang, Z. L., D. H. Beezhold, C. D. Personius, Z. L. Shen.
1993
. Fibronectin cell-binding domain triggered transmembrane signal transduction in human monocytes.
J. Leukocyte Biol.
53
:
79
.
31
Green, D. M., J. Trial, H. H. Birdsall.
1998
. TNF-α, released by co-migrating monocytes, promotes transendothelial migration of activated lymphocytes.
J. Immunol.
161
:
2481
.
32
Gopalan, P. K., A. R. Burns, S. I. Simon, S. Sparks, L. V. McIntire, C. W. Smith.
2000
. Preferential sites for stationary adhesion of neutrophils to cytokine-stimulated HUVEC under flow conditions.
J. Leukocyte Biol.
68
:
47
.
33
Korom, S., W. W. Hancock, A. J. Coito, J. W. Kupiec-Weglinski.
1998
. Blockade of very late antigen-4 integrin binding to fibronectin in allograft recipients. II. Treatment with connecting segment-1 peptides prevents chronic rejection by attenuating arteriosclerotic development and suppressing intragraft T cell and macrophage activation.
Transplantation
65
:
854
.
34
Abe, Y., C. Ballantyne, C. W. Smith.
1996
. Functions of domain 1 and 4 of vascular cell adhesion molecule 1 in α4 integrin-dependent adhesion under static and flow conditions are differentially regulated.
J. Immunol.
157
:
5061
.
35
Konstantopuolos, K., S. Kukreti, C. W. Smith, L. McIntire.
1997
. Endothelial P-selectin and VCAM-1 each can function as primary adhesive mechanisms for T cells under conditions of flow.
J. Leukocyte Biol.
61
:
179
.
36
Alon, R., R. C. Fuhlbrigge, E. B. Finger, T. A. Springer.
1996
. Interactions through L-selectin between leukocytes and adherent leukocytes nucleate rolling adhesions on selectins and VCAM-1 in shear flow.
J. Cell Biol.
135
:
849
.
37
Lim, Y. C., K. Snapp, G. S. Kansas, R. Camphausen, H. Ding, F. W. Luscinskas.
1998
. Important contributions of P-selectin glycoprotein ligand-1-mediated secondary capture to human monocyte adhesion to P-selectin, E-selectin, and TNF-α-activated endothelium under flow in vitro.
J. Immunol.
161
:
2501
.
38
Trial, J., J. A. Rubio, H. H. Birdsall, R. D. Rossen.
2004
. Monocyte activation by circulating fibronectin fragments in HIV-1 infected patients.
J. Immunol.
173
:
2190
.
39
von Adrian, U. H., C. R. Mackay.
2000
. T-cell function and migration. Two sides of the same coin.
N. Engl. J. Med.
343
:
1020
.
40
Randolph, G. J., S. Beaulieu, S. Lebecque, R. M. Steinman, W. A. Muller.
1998
. Differentiation of monocytes into dendritic cells in a model of transendothelial trafficking.
Science
282
:
480
.
41
Weeks, B. S..
1998
. The role of HIV-1 activated leukocyte adhesion mechanisms and matrix metalloproteinase secretion in AIDS pathogenesis.
Int J. Mol. Med.
1
:
361
. (Rev.).
42
Dhawan, S., R. K. Puri, A. Kumar, H. Duplan, J. M. Masson, B. B. Aggarwal.
1997
. Human immunodeficiency virus-1-tat protein induces the cell surface expression of endothelial leukocyte adhesion molecule-1, vascular cell adhesion molecule-1, and intercellular adhesion molecule-1 in human endothelial cells.
Blood
90
:
1535
.
43
Le Naour, R., C. Lussiez, H. Raoul, A. Mabondzo, D. Dormont.
1997
. Expression of cell adhesion molecules at the surface of in vitro human immunodeficiency virus type 1-infected human monocytes: relationships with tumor necrosis factor α, interleukin 1α, and interleukin 6 syntheses.
AIDS Res. Hum. Retroviruses
13
:
841
.
44
Randolph, G. J., G. Sanchez-Schmitz, R. M. Liebman, K. Schakel.
2002
. TheCD16+ (FcγRIII+) subset of human monocytes preferentially becomes migratory dendritic cells in a model tissue setting.
J. Exp. Med.
196
:
517
.
45
Chomarat, P., C. Dantin, L. Bennett, J. Banchereau, A. K. Palucka.
2003
. TNF skews monocyte differentiation from macrophages to dendritic cells.
J. Immunol.
171
:
2262
.
46
Engering, A., S. J. Van Vliet, T. B. Geijtenbeek, Y. Van Kooyk.
2002
. Subset of DC-SIGN+ dendritic cells in human blood transmits HIV-1 to T lymphocytes.
Blood
100
:
1780
.
47
Marshall, D. C., T. Wyss-Coray, C. R. Abraham.
1998
. Induction of matrix metalloproteinase-2 in human immunodeficiency virus-1 glycoprotein 120 transgenic mouse brains.
Neurosci. Lett.
254
:
97
.
48
Shoeman, R., R. Hartig, C. Hauses, P. Traub.
2002
. Organization of focal adhesion plaques is disrupted by action of the HIV-1 protease.
Cell Biol. Int.
26
:
529
.
49
Oswald, M., K. von der Helm.
1991
. Fibronectin is a non-viral substrate for the HIV proteinase.
FEBS Lett.
292
:
298
.