Published online before print July 22, 2003
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Experimental Transplantation and Immunology Branch and
HIV and AIDS Malignancy Branch, National Cancer Institute, National Institutes of Health, Bethesda, Maryland; and
Division of Biostatistics, Center for Biologics Evaluation and Research, U. S. Food and Drug Administration, Rockville, Maryland
1Correspondence: Experimental Transplantation and Immunology Branch, Center for Cancer Research, National Cancer Institute, Building 10, Room 12N226, MSC 1907, Bethesda, MD 20892. E-mail: tosatog{at}mail.nih.gov
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: chemokine • cytokine • chemokine processing • AIDS • anti-HIV therapy
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
CXCR-4 serves as a coreceptor with CD4 for T cell tropic isolates of human immunodeficiency virus type-1 (HIV-1) [27 28 29 30 ]. In vitro, SDF-1 can effectively block T-tropic HIV-1 entry into CD4-positive cells by binding to CXCR-4 expressed by lymphocytes and inducing its internalization [28 , 29 , 31 ]. Polymorphism in a conserved part of the 3' untranslated region of the SDF-1 gene has been found to correlate with delayed onset of AIDS in HIV-infected individuals [32 ]. It was hypothesized that this protective effect of the SDF-1 gene variant was a result of increased levels of SDF-1 protein available to bind to CXCR-4. However, other studies have reported an association between this SDF-1 gene variant and HIV disease acceleration [33 , 34 ]. In addition, studies that examined the relationship between levels of circulating SDF-1 and HIV disease progression have reached contrasting conclusions [35 36 37 ].
The HIV-1-suppressive effects of SDF-1 in vitro contrast with the natural history of HIV-1 infection in humans, suggesting that SDF-1 is not effective at preventing HIV-1 disease progression. To reconcile this apparent discrepancy, we sought to characterize SDF-1 as it is found in vivo.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
SDF-1 enzyme-linked immunosorbent assay (ELISA)
ELISA for SDF-1 was performed as described previously [26
], except that for coating, we combined two mouse monoclonal anti-SDF-1 antibodies (MAB350, clone 79018.11, and MAB310, clone 79014.111, each at 5 µg/ml; R&D Systems, Minneapolis, MN; Table 1 ). rhSDF-1
standard (11–24,300 pg/mL; R&D Systems) or test sera [1:10 and higher dilution in phosphate-buffered saline (PBS) containing 0.05% Tween-20] were added. BAF310 (200 ng/ml; R&D Systems) was used as a secondary antibody, followed by streptavidin horseradish peroxidase (1:200 dilution; R&D Systems). The peroxidase activity was visualized by tetramethoxybenzene peroxidase substrate solution (KPL, Gaithersburg, MD), followed by 2 N H2SO4. Plates were read at 450 nm with correction at 630 nm. The concentration of SDF-1 in test samples was calculated from absorbance values in relation to the standard curve using SOFTmax PRO software. The assay was linear between 11 and 2700 pg/mL SDF-1 standard and had a lower limit of sensitivity calculated at 
33 pg/mL. No difference in assay sensitivity and linearity was observed when the standard SDF-1 preparation was diluted in PBS buffer containing 0.05% Tween-20 alone, 10% fetal calf serum (FCS), or 10% fresh or heat-inactivated serum. The assay was reproducible and specific for SDF-1, as it did not detect human interferon-inducible protein 10 (IP-10), human Exodus-2, human macrophage chemotactic and activating factor, human granulocyte macrophage-colony stimulating factor (GM-CSF), and human interleukin-6 (IL-6) at concentrations ranging between 100 pg and 10 ng/mL.
|
View this table: [in a new window] |
Table 1. Anti SDF-1 Antibodies Used in This Study
|
, obtained from R&D Systems (350-NS), was used throughout. A synthetic form SDF- lacking the C-terminal lysine (amino acids 1–67) was purchased from Upstate Biotechnology (Lake Placid, NY). The specific elastase inhibitor M-methoxysuccinyl-alanine-alanine-proline-proline-valine-chloromethyl ketone was obtained from Calbiochem-Novabiochem Co. (San Diego, CA); the synthetic broad-spectrum metalloproteinase (MMP) inhibitor BB94, British Biotech, was a kind gift of Dr. William Stetler-Stevenson, NCI; the synthetic CD26/dipeptidyl peptidase IV inhibitor peptide Ile-Pro-Ile was a kind gift of Dr. Michael Norcross, Center for Biologics (Bethesda, MD); and the protease inhibitors phenylmethylsulfonyl fluoride (PMSF) and aprotinin were from Sigma Chemical Co. (St Louis, MO). For immunoprecipitation of SDF-1 from human serum (no SDF-1 added to serum), the human serum was precleared by three rounds incubation with protein-G Sepharose four Fast Flow beads (Amersham Biosciences, Piscataway, NJ). Immunoprecipitation of SDF-1 from human serum (fresh or after heating at 56°C for 30 min) with or without addition of exogenous rSDF-1 (R&D Systems, 350-NS) used MAB310, clone 79014.111 (10 µg/mL; R&D Systems), followed by addition of protein-G beads to the antibody/protein mixture. Human IL-10 (R&D Systems) was immunoprecipitated with mouse anti-human IL-10-purified mAb (clone JES3-9D7, PharMingen, San Diego, CA, 18551A). Protein complexes, dissociated from beads by SDS sample and boiling, were run through 10–20% or 16% tricine gels (Invitrogen, Carlsbad, CA) and were blotted onto Immobilon-P membranes (Millipore, Bedford, MA). Rabbit anti-human SDF-1 antigen affinity-purified polyclonal antibody (PeproTech) or BAF310 antigen affinity-purified antibody (R&D Systems) was used for detection of SDF-1. Rabbit IgG anti-human IL-10 antibody (Santa Cruz Biotechnology, Santa Cruz, CA) was used for detection of human IL-10. Bound rabbit or goat antibody was detected with affinity-purified, peroxidase-linked, donkey anti-rabbit or anti-goat IgG antibody (Amersham Biosciences) followed by an enhanced chemiluminescence kit detection system (Amersham Biosciences).
Endothelial cell cultures and migration assays
Human umbilical vein endothelial cells (HUVECs), prepared from umbilical cord as described previously [26
], were propagated through passage four in M199 (Gibco-BRL, Grand Island, NY) culture medium containing 20% newborn calf serum (Sigma Chemical Co.), 5% human AB serum, 1.6 mM L-glutamine (Invitrogen), 50 mg/ml porcine heparin (Sigma Chemical Co.), 50 µg/ml ascorbate (Fisher Scientific, Fail Lawn, NJ), 15 mM HEPES buffer (Calbiochem-Behring, La Jolla, CA), and 15 µg/ml endothelial cell growth supplement (a crude extract of bovine neural tissue containing basic and acidic fibroblast growth factors, Sigma Chemical Co.). HUVECs were detached with 2 mM EDTA in PBS, washed, and suspended in complete culture medium.
Endothelial cell migration assays were performed as described previously [38
] using Costar transwells (6.5 µm in diameter, 8 µm/pore, Corning Costar, Cambridge, MA). Treatment of SDF-1 with human serum (fresh or heated at 56°C for 30 min) was performed by incubating SDF-1 (1 µg in 20 µl PBS) with 10% human serum (2 µl serum added to 1 µg SDF-1 in 20 µl PBS) for 10 min at room temperature. rSDF-1 (100 ng, R&D Systems), alone or exposed to human serum as described above, was added to 0.5 ml medium [RPMI-1640 medium with 0.5% bovine serum albumin (BSA) and 10 nM HEPES] to the bottom chamber, and 1 x 106 endothelial cells (HUVEC, passage three) were added to the upper chamber in RPMI-1640 medium with 0.5% BSA and 10 nM HEPES. After incubation for 4 h at 37°C, the number of migrated cells was counted.
Cell proliferation assays
The murine DW34 cells [39
] (a gift of Dr. Paul Kincade, Oklahoma Medical Research Foundation, Oklahoma City) were maintained on the murine stromal cell line MS-5 [40
] (a gift of Dr. Anna Berardi, Ospedale Bambin Gesu, Rome, Italy) in RPMI-1640 medium supplemented with 50 µM 2-mercaptoethanol (2-ME) and 10% FCS. DW34 were harvested, washed, and cultured at 2 x 104 cells/well in RPMI-1640 medium containing 10% FCS and 50 µM 2-ME in 96-well flat-bottom plates for 30 h at 37°C. SDF-1 (R&D Systems), untreated or exposed to 10% human serum as described above, was added to the cultures at final concentrations ranging between 100 and 0.8 ng/ml. All assays were performed in triplicate. The cultures were pulsed with 0.5 µCi [40
] thymidine (Perkin-Elmer, Foster City, CA) during the last 6 h of culture, the cells harvested, and radioactivity measured by liquid scintillation counting.
Flow cytometry
HUVECs (passage four) were detached from plates with 2 mM EDTA in PBS, washed in ice-cold binding buffer (RPMI-1640 medium with 20 mM HEPES and 1% BSA), and incubated with mouse IgG1 isotype control (1 µg/ml) for 30 min. Subsequently, the cells (5x105 cells/ml) were incubated (90 min at 4°C) in buffer alone (0.1 ml PBS with 0.1% BSA), buffer containing 10% fresh or heat-treated (56°C for 30 min) human serum, or SDF-1 (1 µg), diluted in buffer alone or exposed to 10% fresh or heat-treated (56°C for 30 min) human serum, as described above. After washing in cold-binding buffer, surface SDF-1 was revealed by clone MAB310 (R&D Systems; 5 µg/ml for 45 min at 4°C) followed by a phycoerythrin-labeled goat anti-mouse F(ab')2 fragment (30 min at 4°C; Jackson ImmunoResearch, West Grove, PA). Data were collected from 5 x 103 viable cells using FACScalibur cytofluorometer (Becton Dickinson, Franklin Lakes, NJ) and were analyzed using CELLQuest software (Becton Dickinson). Background fluorescence was assessed through staining with isotype-matched antibodies.
HIV-1 infection of peripheral blood mononuclear cells (PBMCs)
PBMCs, obtained from healthy, normal blood donors, were stimulated (1x106 cells/ml) for 48 h with phytohemagglutinin (PHA; Invitrogen) in RPMI-1640 medium supplemented with 10% FCS (Biofluids, Inc., Rockville, MD). The cells (4x106 cells/condition in 4 ml culture medium) were then infected for 1.5 h with 20 ng HIV-1 p24 obtained from H9 cells chronically infected with HIV-1NL4-3 in the presence of SDF-1 (100 nM). The SDF-1 had been incubated with 10% human serum, fresh or heat-treated (56°C for 30 min), as described above. The cells were subsequently washed (four times with PBS) and then suspended in culture medium (RPMI 1640 with 10% FCS) supplemented with IL-2 (20 ng/ml; PeproTech) and rSDF-1 (treated with 10% fresh or heat-inactivated human serum). Cultures were re-fed every 3 or 4 days with culture medium containing IL-2 (20 ng/ml) and SDF-1 (suspended in PBS containing 10% fresh or heat-inactivated human serum). HIV-1 p24 was measured in the culture supernatant by radioimmunoassay (Dupont, Wilmington, DE).
Statistical analysis
The means and 95% confidence intervals were calculated by conventional formulas. Statistical significance of group differences was calculated by Student’s t-test and Dunnett’s method at 0.05 level of significance, with correction for multiple comparisons [41
]. Correlation coefficients were calculated using Sperman’s Rho test.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
 View larger version (17K): [in a new window] |
Figure 1. SDF-1 levels in serum samples from blood donors and AIDS patients with or without KS. Data points reflect serum SDF-1 levels measured in individual blood donors (n=39) and AIDS (HIV pos) patients (n=73). Each point reflects the means of triplicate determinations. Horizontal lines denote the mean SDF-1 group levels.
|
diluted in buffer alone. We therefore tested the effects of human serum on our ability to identify rSDF-1 by this method. Purified rSDF-1 (100 ng) was incubated with fresh or heat-treated (56°C for 30 min) human serum, immunoprecipitated, and immunoblotted with rabbit anti-SDF-1 antibodies. A band attributable to SDF-1 was identified when SDF-1 was incubated for 10 min with heat-treated human serum (Fig. 2A
, lane 4). However, when SDF-1 was incubated with fresh human serum under otherwise identical conditions, only a very faint band attributable to SDF-1 was visualized (Fig. 2A
, lane 3). No lower or higher molecular weight bands attributable to SDF-1 were noted. As expected, immunoprecipitation of human serum alone, fresh (Fig. 2A
, lane 1) or heat-treated (Fig. 2A
, lane 2), did not yield SDF-1-related bands. In contrast to the results with SDF-1, IL-10-related bands were similarly detected after immunoprecipitation and immunoblotting of IL-10 incubated with the same fresh or heated serum (Fig. 2B)
. In additional experiments, rSDF-1 (30 ng) was diluted into fresh or heated (56°C for 30 min) human serum and directly immunoblotted (using rabbit anti-SDF-1 antibody). Again, a clear band attributable to SDF-1 was visualized when SDF-1 was diluted in heat-treated human serum, whereas no SDF-1-related band was visualized when the chemokine was diluted in fresh human serum (Fig. 2C)
. The absence of an SDF-1-related band from the sample containing fresh serum could not be attributed to uneven loading, as equal amounts of protein were loaded. In addition, when tested in parallel by ELISA (using mouse monoclonal anti-SDF-1 antibodies for coating and goat anti-human SDF-1 for detection), SDF-1 concentrations were found to be comparable when rSDF-1 (30 ng/ml) was diluted in fresh or heat-treated human serum (Fig. 2D)
. These results suggested that SDF-1 is masked or degraded upon exposure to fresh serum such that it can no longer be identified by immunoblotting under the conditions used.
 View larger version (47K): [in a new window] |
Figure 2. Effects of human serum on SDF-1 detection by immunoblotting. (A) SDF-1 immunoprecipitation and immunoblotting. rSDF-1 (100 ng) in 1 ml M199 medium containing 10% precleared human serum, fresh or heat-treated (56°C, 30 min), was immunoprecipitated with mouse monoclonal SDF-1 antibody. Control medium containing fresh or heated serum was immunoprecipitated in parallel. The immunoprecipitates were immunoblotted with rabbit antibodies against SDF-1. Lane 1, Fresh human serum; lane 2, heat-treated human serum; lane 3, SDF-1 in fresh human serum; lane 4, SDF-1 in heat-treated human serum. (B) IL-10 immunoprecipitation and immunoblotting. Experimental conditions identical to those described in A with the exception that IL-10 (100 ng) was immunoprecipitated with mouse monoclonal IL-10 antibodies and immunoblotted with rabbit antibodies against IL-10. Lane 1, Fresh human serum; lane 2, heat-treated human serum; lane 3, IL-10 in fresh human serum; lane 4, IL-10 in heat-treated human serum. (C) SDF-1 immunoblotting. rSDF-1 (30 ng) in 10 µl M199 medium containing 10% precleared human serum, fresh or heat-treated (56°C, 30 min), was run through a 10–20% tricine gel and immunoblotted with rabbit antibodies against SDF-1. Lane 1, Fresh human serum; lane 2, heat-treated human serum; lane 3, SDF-1 in fresh human serum; lane 4, SDF-1 in heat-treated human serum. (D) SDF-1 concentration measured by ELISA. rSDF-1 (30 ng) was diluted into 1 ml M199 medium containing 10% precleared human serum, fresh or heat-treated (56°C, 30 min). SDF-1 concentrations were measured in medium alone (containing 10% precleared serum, fresh or heat-treated) or medium supplemented with SDF-1.
|
(1 µg) in buffer alone or with 10% heat-treated (56°C for 30 min) human serum (Fig. 3A
and 3C
, respectively). However, no such increase of surface SDF-1 was derived from rSDF-1 in 10% fresh human serum under otherwise identical conditions (Fig. 3B)
. Thus, SDF-1 binding to endothelial cells is not detectable by FACS when the chemokine is exposed to fresh serum.
 View larger version (28K): [in a new window] |
Figure 3. Effects of human serum on SDF-1 biological activities. (A–C) Fluorescein-activated cell sorter (FACS) analysis of SDF-1 binding to endothelial cells. HUVECs were incubated at 4°C for 90 min in buffer alone or in buffer containing rSDF-1 (1 µg) diluted in buffer alone or in buffer containing 10% fresh or heat-treated human serum. After washing, surface SDF-1 was assessed after staining HUVECs with MAB350 or mouse IgG1 isotype control. (A) Surface SDF-1 in HUVECs incubated in buffer alone (none) and after incubation with SDF-1. (B) Surface SDF-1 in HUVECs incubated in buffer alone (none) and after incubation with SDF-1 diluted in fresh human serum. (C) Surface SDF-1 in HUVECs incubated in buffer alone (none) and after incubation with SDF-1 diluted in heat-treated human serum. (D) SDF-1-induced endothelial cell migration. Chemotactic response of HUVECs to SDF-1 (100 ng/ml) diluted in buffer alone or pretreated with 10% human serum, fresh or heat-treated. The results reflect the mean (±SD; three experiments) number of cells migrated. (E) Proliferation of the pre-B cell clone SW34 in the presence of SDF-1. SW34 cells were cultured for 30 h in medium alone or with SDF-1 (0.8–100 ng/ml) that had been diluted in buffer alone or pretreated with 10% human serum, fresh or heated. Representative experiment of five performed. The results reflect the mean cpm/culture of triplicate cultures. (F) HIV-1 infection of PBMCs in the presence of SDF-1 and human serum. PBMCs (1x106 cell/ml, 4 ml total culture volume) were first activated with PHA (1 µg/ml, 48 h) and then infected with 20 ng HIV-1NL4-3 in the presence of SDF-1 (100 nM) pretreated with 10% fresh or heat-treated human serum. After 1.5 h incubation, the cells were washed, and incubation continued in culture medium (4 ml) supplemented with 20 ng/ml IL-2 and SDF-1 (100 nM) pretreated with 10% fresh or heated serum. HIV-1 p24 was measured in the culture supernatant at the indicated time-points. The results reflect the means of duplicate determinations.
|
treated with 10% heated (56°C for 30 min) human serum was active as endothelial cell chemoattractant across transwells. However, rSDF-1 treated with 10% fresh human serum was inactive in this assay, providing additional evidence that once exposed to fresh serum, SDF-1 is functionally compromised.
The murine pre-B cell clone DW34 responds to SDF-1 with increased proliferation [43
]. We examined the effects of serum on rSDF-1 stimulation of DW34 cell proliferation (Fig. 3E)
. rSDF-1 pretreated with 10% heated (56°C for 30 min) human serum stimulated DW34 cell proliferation to a similar degree as did SDF-1 diluted in buffer alone. However, rSDF-1 pretreated with 10% fresh human serum was only minimally stimulatory at the highest concentration. Thus, exposure to fresh human serum markedly reduces SDF-1 ability to stimulate proliferation of SDF-1-responsive cells.
As SDF-1 can selectively inhibit lymphocyte infection by T cell tropic HIV-1 in vitro [27
28
29
], we examined the effects of serum on this biological activity of SDF-1. To this end, mononuclear cells were first activated with PHA and were subsequently exposed to HIV-1NL4-3 in the presence of 100 nM SDF-1 that had been pretreated with fresh or heated normal human serum. HIV-1 p24 accumulation in the culture supernatant was measured after 3, 7, 10, and 12 days. As shown (Fig. 3E)
, levels of HIV-1 p24 were significantly greater in cultures containing rSDF-1 treated with fresh serum as opposed to cultures containing rSDF-1 treated with heated serum. Thus, fresh serum impairs the anti-HIV activity of SDF-1. Together, these functional experiments provide evidence that fresh serum modifies SDF-1, resulting in a functionally inactive molecule.
Characterization of SDF-1 modification by serum
To investigate the nature of SDF-1 modification by serum and reconcile the differences between SDF-1 detection by ELISA and immunoblotting, we tested the goat antibody (BAF310, R&D Systems) used successfully in the ELISA for SDF-1 capture for its ability to recognize SDF-1 in immunobloting experiments. As shown (Fig. 4A
), this antibody identified rSDF-1 that had been diluted in fresh or heat-treated (56°C for 30 min) human serum equally well. Noteworthy, however, the relative molecular weight of the SDF-1-related band from fresh serum (lanes 1 and 3) was slightly smaller than that from heated serum (lane 2). This apparent size reduction, confirmed in five separate experiments, suggested that serum contains a heat-labile activity that can cleave a portion of the SDF-1 molecule or otherwise modify it to reduce its apparent size. A time course experiment revealed that this effect of serum (10%) was complete within 1–2 min incubation of rSDF-1 (30 ng) at room temperature such that the chemokine could no longer be detected by the rabbit anti-SDF-1 antiserum (not shown).
 View larger version (39K): [in a new window] |
Figure 4. SDF-1 cleavage by serum and effects of inhibitors. (A) SDF-1 (30 ng) in 10 µl M199 medium containing 10% precleared human serum, fresh (F) or heat-treated (H; 56°C, 30 min), was subjected to SDS-PAGE and immunoblotted with goat antibodies against SDF-1. Lane 1, SDF-1 in fresh human serum; lane 2, SDF-1 in heat-treated human serum; lane 3, SDF-1 diluted in fresh human serum and subsequently heat-treated. (B) SDF-1 (50, 25, 12.5, and 6 ng in 1 ml M199 medium containing 1% BSA) and normal human serum (precleared, 2 ml) were immunoprecipitated with MAB310 (10 µg/ml, R&D Systems). After SDS-PAGE and transfer, the membrane was first immunoblotted with rabbit anti-SDF-1 Ab (PeproTech) and then reblotted with BAF310 (R&D Systems). Lane 1, SDF-1 50 ng; lane 2, SDF-1 25 ng; lane 3, SDF-1 12.5 ng; lane 4, SDF-1 6 ng; lane 5, human serum (2 ml). (C) Fresh or heat-treated human serum (30 µl) was incubated alone, with EDTA (5, 10, and 20 mM), or with BB94 (20 µM) for 2 h at 37°C and was then spiked with SDF-1 (150 ng) and incubated at room temperature for 15 min. An aliquot of the mixture was subjected to SDS-PAGE and immunoblotted with rabbit antibodies against SDF-1 (upper panel). The membrane was stripped and reblotted with goat anti-SDF-1 antibodies (lower panel). Lane 1, SDF-1 in fresh human serum; lane 2, SDF-1 in heat-treated human serum; lane 3, SDF-1 diluted in fresh human serum with 5 mM EDTA; lane 4, SDF-1 diluted in fresh human serum with 10 mM EDTA; lane 5, SDF-1 diluted in fresh human serum with 20 mM EDTA; lane 6, SDF-1 diluted in fresh human serum with 20 µM BB94. (D) Fresh or heat-treated human serum (30 µl) was incubated alone, with dipeptidyl peptidase IV/CD26 inhibitor (Ile-Pro-Ile), or with the leukocyte elastase inhibitor N-methoxysuccinil-Ala-Ala-Pro-Val-chloromethyl ketone (MS-AAPV-CMK) for 2 h at 37°C and was then spiked with SDF-1 (150 ng) and incubated at room temperature for 15 min. An aliquot of the mixture was subjected to SDS-PAGE and immunoblotted with rabbit antibodies against SDF-1 (upper panel). The membrane was stripped and reblotted with goat anti-SDF-1 antibodies (lower panel). Lane 1, SDF-1 in fresh human serum; lane 2, SDF-1 in heat-treated human serum; lane 3, SDF-1 diluted in fresh human serum with Ile-Pro-Ile (100 µM); lane 4, SDF-1 diluted in fresh human serum with Ile-Pro-Ile (200 µM); lane 5, SDF-1 diluted in fresh human serum with MS-AAPV-CMK (2.5 µM); lane 6, SDF-1 diluted in fresh human serum with MS-AAPV-CMK (5.0 µM). (E) rSDF-1 (amino acids 1–68) and synthetic SDF-1 (amino acids 1–67) were incubated for 15 min at room temperature in PBS alone or with 10% human serum, fresh or heat-treated (56°C, 30 min). The samples were subjected to SDS-PAGE and immunoblotted with rabbit antibodies against SDF-1 (upper panel). The membrane was stripped and reblotted with goat anti-SDF-1 antibodies (lower panel). Lane 1, rSDF-1; lane 2, rSDF-1 incubated in 10% fresh serum; lane 3, rSDF-1 incubated in 10% heated serum; lane 4, synthetic SDF-1; lane 5, synthetic SDF-1 incubated in 10% fresh serum; lane 6, synthetic SDF-1 incubated in 10% heated serum. (F) Proliferation of SW34 pre-B cells in the presence of full-length SDF-1 and SDF-1 lacking the C-terminal lysine. SW34 cells were cultured for 30 h in medium supplemented with SDF-1 (amino acids 1–68) or SDF-1 (amino acids 1–67) at 100, 50, 25, and 12.5 ng/ml. Representative experiment of four was performed. The results reflect the mean cpm/culture of triplicate determinations. Solid bars, SDF-1 amino acids 1–67; shaded bars, SDF-1 1–68. SW34 cell proliferation in medium alone (no SDF-1) was 2590 (±356) cpm/culture.
|
Previous experiments have shown that SDF-1 can be enzymatically cleaved in vitro by MMPs 1, 3, 9, 13, and 14, CD26/dipeptidyl peptidase IV, serine proteases, and leukocyte elastase to generate distinct N-terminally truncated forms of the molecule [44
45
46
47
]. To assess whether these proteases are responsible for serum SDF-1 cleavage, several specific inhibitors were used in the presence of rSDF-1 and serum. The broad-spectrum, synthetic MMP inhibitor BB-94 (20 µM) [48
] (Fig. 4C)
and the leukocyte elastase inhibitor [47
] N-methoxysuccinyl-alanine-alanine-proline-valine-chloromethyl ketone (2.5 and 5 µM; Fig. 4D
) did not prevent rSDF-1 modification by serum. By contrast, the metal chelator EDTA (at 5, 10, and 20 mM) prevented the apparent SDF-1 size reduction by serum (Fig. 4C)
. The CD26/dipeptidyl peptidase IV competitive inhibitors [44
, 49
] Ile-Pro-Ile (100 and 200 µM; Fig. 4D ) and IP-10 (300 µM; not shown), the irreversible CD26/dipeptidyl peptidase IV inhibitor AB192 (25 µM, 15 min incubation at 37°C; not shown) [50
], and the serine protease inhibitor PMSF (1 mM; not shown) failed to prevent SDF-1 modification by serum. These results demonstrate that serum modification of SDF-1 is metal-dependent. The metal dependence of SDF-1 cleavage by serum was confirmed by using the metal chelator cis, cis-1,3,5-triaminocyclohexane (tachpiridine, 10 mM) [51
], which also protected SDF-1 (results not shown). In addition, these results provide evidence that serum alteration of SDF-1 is not attributable to the effects of MMPs, CD26/dipeptidyl peptidase IV, leukocyte elastase, or serine proteases, which can truncate SDF-1 at the N-terminus.
Further evidence that serum modification of SDF-1 differs from cleavage induced by these enzymes was obtained using a synthetic SDF-1 molecule, which lacks the C-terminal lysine (Upstate Biotechnology). As shown in Figure 4E
, the anti-SDF-1 rabbit antiserum (PeproTech) failed to recognize this synthetic SDF-1 molecule (lanes 3–6), but BAF310 (R&D Systems) did. This pattern of antibody recognition of synthetic SDF-1 mirrors the situation with full-length SDF-1 exposed to fresh human serum (lane 2), providing evidence that serum modification of SDF-1 likely involves the C-terminal portion of the molecule. We compared the activities of full-length SDF-1 (1–68; R&D Systems) and synthetic SDF-1 (1–67; Upstate Biotechnology) as stimulators of DW34 pre-B cell proliferation in vitro. As shown (Fig. 4F)
, both molecules dose-dependently stimulated cell proliferation, but full-length SDF-1 (1–68) was more active than the SDF-1 molecule, which lacks the C-terminal lysine (1–67). These results provide evidence that truncations at the C-terminal end can reduce the biological activity of SDF-1.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
is rapidly altered to yield a molecule that appears smaller and is biologically inactive as a chemoattractant for CXCR-4-expressing cells and as a pre-B cell growth factor. It is also inactive as an inhibitor of T cell tropic HIV-1 infection. The rapid nature of SDF-1 modification by serum and the absence of detectable, native SDF-1 in serum provide strong evidence that circulating SDF-1 is largely inactive and would not be expected to block the infection of cells by T cell tropic HIV-1 present in blood. SDF-1 is emerging as a primary regulator of cell migration to specific tissue sites [5 ]. The homing and retention of hematopoietic stem cells to the bone marrow (BM) [19 , 20 ], plasma cell movement to secondary organs [52 ], and cancer cell metastasis to the liver, BM, and lymph nodes [25 ] appear to depend on the interaction between CXCR-4 expressed by the migrating cells and SDF-1 in the homing tissue. Recently, stem-cell mobilization induced by the administration of G-CSF was attributed to a decrease of BM SDF-1 [21 , 53 ]. All these biological activities of SDF-1 are dependent on local chemokine gradients, likely derived from SDF-1 produced locally. This feature distinguishes SDF-1 from many cytokines that display prominent, systemic effects. Given this paradigm, serum modification of SDF-1 could serve to prevent systemic accumulation of SDF-1, which would blunt the effects of local gradients. As SDF-1 is constitutively expressed by a variety of cells, degradation is likely to serve a critical, regulatory function for this chemokine.
Recently, the possibility that systemic SDF-1 might be used therapeutically has been examined. Treatment with sulfated polysaccharides and injection of an adenoviral vector expressing SDF-1 were reported to increase levels of circulating SDF-1 and to induce a transient mobilization of stem/progenitor cells to the peripheral blood [54 , 55 ]. However, we would predict that administration of SDF-1 or drugs that promote endogenous SDF-1 production will not induce sustained change in levels of biologically active SDF-1 in the circulation as a result of its rapid degradation. Rather, drugs that block SDF-1 cleavage activity in serum of modified forms of SDF-1, which are resistant to cleavage, could be useful at increasing systemic levels of biologically active SDF-1.
Based on the fact that the serum-modified molecule differs in relative size from the full-length molecule only slightly, we hypothesize that 10 or fewer amino acids are removed. Mass spectrometry analysis could not identify SDF-1 fragments generated after exposure to serum as a result of the confounding effects of serum components. However, the metal chelators EDTA and tachpiridine protected rSDF-1 from serum cleavage, suggesting a metal-dependent reaction. It should be noted that certain enzymes have been reported to cleave SDF-1 under experimental conditions. CD26/dipeptidyl peptidase IV, which is expressed as a surface molecule by activated T lymphocytes and other cells and is present in plasma in a catalytically active soluble form, was previously reported to cleave SDF-1 in vitro to generate a truncated and inactive chemokine lacking the amino-terminal two amino acids [44
]. The serine protease cathepsin G, which is expressed on the cell surface by B lymphocytes, natural killer cells, and other cells and is released by neutrophils and monocytes, was also found to cleave SDF-1 in vitro to generate an inactive chemokine lacking the five N-terminal amino acids [46
]. In addition, leukocyte elastase, a protease released by mononuclear blood cells and neutrophils, generated an SDF-1 fragment in vitro, which is lacking the three amino-terminal residues and is inactive [47
]. Finally, the matrix MMP gelatinase-A and other MMPs were found to selectively cleave and inactivate SDF-1 in vitro to release an N-terminal tetrapeptide [45
]. However, specific inhibitors of CD26/dipeptidyl peptidase IV, serine proteases, leukocyte elastase, and matrix MMP did not block serum modification of SDF-1, providing evidence that these proteases are not responsible for serum inactivation. In addition, we have evidence that serum modification of SDF-1 involves the C-terminal portion of the molecule, as an antibody that fails to recognize a synthetic form of SDF-1, which lacks the C-terminal lysine, also fails to recognize serum-modified SDF-1. Thus, although selected proteases—some present in the circulation—can specifically cleave SDF-1 at the N-terminous under certain conditions, these do not appear to be responsible for its inactivation in serum, and the enzyme that modifies SDF-1 remains to be defined. Additional studies will be directed at identifying this activity and defining its mechanisms of action, but the observation that serum SDF-1 is modified and inactive has profound implications on the chemokine range of biological activities.
Received April 11, 2003; revised June 10, 2003; accepted June 11, 2003.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  F. Vianello, N. Papeta, T. Chen, P. Kraft, N. White, W. K. Hart, M. F. Kircher, E. Swart, S. Rhee, G. Palu, et al. Murine B16 Melanomas Expressing High Levels of the Chemokine Stromal-Derived Factor-1/CXCL12 Induce Tumor-Specific T Cell Chemorepulsion and Escape from Immune Control. J. Immunol., March 1, 2006; 176(5): 2902 - 2914. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Dunleavy, F. Hakim, H. K. Kim, J. E. Janik, N. Grant, T. Nakayama, T. White, G. Wright, L. Kwak, R. Gress, et al. B-cell recovery following rituximab-based therapy is associated with perturbations in stromal derived factor-1 and granulocyte homeostasis Blood, August 1, 2005; 106(3): 795 - 802. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. A. Davis, K. E. Singer, M. De La Luz Sierra, M. Narazaki, F. Yang, H. M. Fales, R. Yarchoan, and G. Tosato Identification of carboxypeptidase N as an enzyme responsible for C-terminal cleavage of stromal cell-derived factor-1{alpha} in the circulation Blood, June 15, 2005; 105(12): 4561 - 4568. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. Salvucci, M. Basik, L. Yao, R. Bianchi, and G. Tosato Evidence for the involvement of SDF-1 and CXCR4 in the disruption of endothelial cell-branching morphogenesis and angiogenesis by TNF-{alpha} and IFN-{gamma} J. Leukoc. Biol., July 1, 2004; 76(1): 217 - 226. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. De La Luz Sierra, F. Yang, M. Narazaki, O. Salvucci, D. Davis, R. Yarchoan, H. H. Zhang, H. Fales, and G. Tosato Differential processing of stromal-derived factor-1{alpha} and stromal-derived factor-1{beta} explains functional diversity Blood, April 1, 2004; 103(7): 2452 - 2459. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
