Chlamydia pneumoniae uses the mannose 6-phosphate/insulin-like growth factor 2 receptor for infection of endothelial cells

Abstract

Several mechanisms for attachment and entry of Chlamydia have been proposed. We previously determined that the major outer membrane protein of Chlamydia trachomatis is glycosylated with a high-mannose oligosaccharide, and a similar structure inhibited the attachment and infectivity of C. trachomatis in epithelial cells. Because insulin-like growth factor 2 (IGF2) was shown to enhance the infectivity of Chlamydia pneumoniae but not C. trachomatis in endothelial cells, a hapten inhibition assay was used to analyze whether the mannose 6-phosphate (M6P)/IGF2 receptor that also binds M6P could be involved in infection of endothelial cells (HMEC-1) by Chlamydia. M6P and mannose 6-phosphate-poly[N-(2-hydroxyethyl)-acrylamide] (M6P-PAA) inhibited the infectivity of C. pneumoniae AR-39, but not C. trachomatis serovar UW5 or L2, while mannan inhibited the growth of C. trachomatis, but not C. pneumoniae. Using metabolically labeled organisms incubated with cells at 4 degrees C (organisms attach but do not enter) or at 37 degrees C (organisms attach and are internalized), M6P-PAA was shown to inhibit attachment and internalization of C. pneumoniae in endothelial cells but did not inhibit attachment or internalization of C. trachomatis serovar E or L2. These findings indicate that C. pneumoniae can utilize the M6P/IGF2 receptor and that the use of this receptor for attachment and entry differs between C. pneumoniae and C. trachomatis.

FIG. 1.

Effects of M6P, mannan, G6P, and heparin, all trans RA (ATRA), and TTNPB on infectivity of human arterial endothelial cells (HMEC-1). The endothelial cells were inoculated with C. pneumoniae (AR-39) or C. trachomatis (serovar E) in the presence of the ligands, and the inoculated cells were cultured for 2 or 3 days. For RA experiments, endothelial cells were incubated for 1 h prior to inoculation with RA or TTNPB and washed with PBS prior to inoculation or the retinoid was added to the inoculum as described above. Infectivity of C. pneumoniae and C. trachomatis is expressed as mean inclusion counts per well (as a percentage of the control). The error bars indicate 1 standard deviation from the means of inclusion counts from triplicate coverslips. *, P < 0.05, and **, P < 0.01 (Student's t test) for culture with the indicated ligands versus culture without a ligand. (A) Inhibition of C. pneumoniae infection by M6P is dose dependent, while G6P and mannan have no effect. (B) Heparin inhibition of C. pneumoniae infection of endothelial cells is dose dependent. (C) RA inhibits C. pneumoniae infectivity of endothelial cells, while its analog, TTNPB, which binds to the RAR, does not.

FIG. 2.

Effects of M6P-PAA treatment on growth of C. pneumoniae and C. trachomatis and attachment to and internalization by endothelial cells. For hapten inhibition studies, endothelial cells were inoculated with C. pneumoniae (AR-39) or C. trachomatis (UW5) in the presence of different concentrations of M6P-PAA. Growth of C. pneumoniae and C. trachomatis is expressed as mean inclusion counts per well (as a percentage of the control). For binding assays, endothelial cells were inoculated with [³⁵S]methionine-labeled purified C. pneumoniae organisms in the presence of M6P-PAA for 2 h at 4°C or 37°C and washed. The radioactivity in cell lysates was counted and expressed as the mean of cpm per well (as a percentage of the control). The error bars indicate 1 standard deviation from the means of cpm from triplicate wells. (A) Effect of M6P-PAA treatment on C. pneumoniae. P < 0.05 (Student's t test) for treated cultures versus untreated cultures (control) at all time points for growth (infectivity), attachment, and internalization of C. pneumoniae. (B) Effect of M6P-PAA treatment on C. trachomatis. There were no statistically significant differences observed in treated cells versus untreated controls.

FIG. 3.

Infectivity of C. pneumoniae in mouse L cells transfected with the bovine M6P/IGF2 receptor. (A) Inhibition of C. pneumoniae (AR-39) inclusion formation by β-glucuronidase in mouse L cells transfected with bovine M6P/IGF2 receptor (CC2) but not in untransfected D9 cells. The mouse L cells (D9) and CC2 cells were treated with β-glucuronidase for 30 min at 4°C and then inoculated with C. pneumoniae in the presence of β-glucuronidase. Infectivity of C. pneumoniae is expressed as the mean of inclusion counts per well or as a percentage of the control. The error bars indicate 1 standard deviation (SD) from the means of inclusion counts from triplicate coverslips. *, P < 0.05 (Student's t test) for treated cultures versus untreated cultures. (B) Infectivity of C. trachomatis L2 for CC2 cells was not inhibited by β-glucuronidase treatment. (C) Inhibition of C. pneumoniae inclusion formation in mouse L cells transfected with bovine M6P/IGF2 receptor (CC2) by wortmannin pretreatment. Cells were treated with wortmannin for 30 min at 37°C and then inoculated with C. pneumoniae in the presence of wortmannin.