Herpes simplex virus-induced epithelial damage and susceptibility to human immunodeficiency virus type 1 infection in human cervical organ culture

Abstract

Normal human premenopausal cervical tissue has been used to derive primary cell populations and to establish ex vivo organ culture systems to study infections with herpes simplex virus (HSV-1 or HSV-2) and human immunodeficiency virus type 1 (HIV-1). Infection with either HSV-1 or HSV-2 rapidly induced multinuclear giant cell formation and widespread damage in mucosal epithelial cells. Subsequent exposure of the damaged mucosal surfaces to HIV-1 revealed frequent co-localization of HSV and HIV-1 antigens. The short-term organ culture system provides direct experimental support for the epidemiological findings that pre-existing sexually transmitted infections, including primary and recurrent herpes virus infections at mucosal surfaces, represent major risk factors for acquisition of primary HIV-1 infection. Epithelial damage in combination with pre-existing inflammation, as described here for overtly normal human premenopausal cervix, creates a highly susceptible environment for the initiation and establishment of primary HIV-1 infection in the sub-mucosa of the cervical transformation zone.

Figure 1. HSV infection of primary human cell populations.

Primary human cervical epithelial cells were propagated on transwell membranes and then infected with 1×10⁵ plaque forming units (pfu) of HSV-1 or HSV-2. Localized disruptions in the epithelial cell monolayers were visible after 17 hours for HSV-2 infections and 24–36 hours for HSV-1 infections. A, B: Phase contrast photomicrographs taken at 36 hours post HSV-2 infection. C: Phase contrast photomicrograph of control uninfected primary cervical epithelial cells. D: Detection of HSV-1 antigens by standard colorimetric immunohistochemistry: mouse monoclonal anti-HSV-gB antibody. Brown stain reveals a focus of HSV-1-infected cells at 21 hours post infection. E, F: Detection of HSV-1 antigens at 36 hours post infection by fluorescence microscopy; red: mixture of mouse anti-HSV-ICP4 and mouse anti-HSV-gB monoclonal antibodies, green: rabbit polyclonal anti-cytokeratin antibody, blue: TOTO-3 nuclear stain. Inset in E shows uninfected primary epithelial cells stained with the same antibody mixture. Note also that the lower right corner of F contains a large region of uninfected cells with normal epithelial morphology. Primary human fibroblasts were propagated on glass chamber slides, infected with varying doses of HSV-2 then fixed and processed for standard immunohistochemical detection with a mouse monoclonal antibody directed against HSV-ICP4. G, H: Brown stain reveals extensive HSV-2 infection at 24 hours post infection. Note the presence of some uninfected cells in the upper left area of G and the extensive morphological changes in the infected cell population with 100-fold higher inoculum of HSV-2 in H. Essentially all of the cells were infected with the 1×10⁵ pfu inoculum used in H. I: Control population of uninfected primary cervical fibroblasts that were fixed and processed in parallel with the anti-HSV-ICP4 antibody. Bars A–C = 250 µm; D = 50 µm; E, F = 100 µm; G–I = 250 µm.

Figure 2. Epithelial damage induced in HSV-infected cervical tissue.

Histopathological changes induced in HSV-2-infected ectocervix include areas of extensive epithelial damage with multinucleated giant cells. A, B: Hematoxylin and eosin (H&E) staining of 5 µm tissue sections; pieces of ectocervix were infected with 1×10⁵ pfu of HSV-2 and fixed after 6 days in organ culture. Arrowheads in A indicate multinucleated giant cells – 7 giant cells are visible in ∼500 µm of epithelial surface shown in the figure. C: Control section from an adjacent piece of ectocervix that was incubated ex vivo for 6 days without exposure to HSV – some degeneration of the stratified squamous epithelium has occurred but no multinucleated giant cells are present. D, E: Independent sections of ectocervix from the HSV-2-infected tissue shown in panels A and B, processed for standard colorimetric immunohistochemical detection of viral antigens. D: Control antibody directed against HIV-1 p24 gag showing no positive signal (brown stain) despite the presence of multinucleated giant cells. E: Rabbit polyclonal anti-HSV antibody showing strong positive signal (brown stain) in multinucleated giant cells at the disrupted epithelial surface. F: Adjacent section from the HSV-2-infected ectocervix shown in A, B and D, E that was analyzed by confocal fluorescence microscopy; green: rabbit polyclonal anti-HSV antibody, blue: TOTO-3 nuclear stain. G: Control, uninfected tissue processed in parallel with F, using the same antibody mix and the same parameters for fluorescent image capture. H: H&E staining of a 5 µm tissue section of HSV-2-infected endocervix, fixed after 7 days in organ culture. Note the extensive areas of epithelial damage and multinucleated giant cells (arrows). Normal columnar endocervical epithelial cells are visible along the upper and lower surfaces of this tissue section. Bars A, B = 100 µm; C = 250 µm; D–H = 100 µm.

Figure 3. Co-localization of viral antigens in multinucleated giant cells generated from cervical tissue dually-infected with HSV-1+HIV-1.

Tissue pieces were infected with 1×10⁵ pfu of HSV-1, incubated for 24 hours then infected with HIV-1 (virus equivalent to 20 pg of p24gag) in normal seminal plasma and then incubated for an additional 120 hours. A, B: Giant cell aggregates visualized by confocal fluorescence microscopy; green: rabbit polyclonal anti-HSV antibody, red: mouse monoclonal anti-HIV-1 p24 gag antibody with TSA enhancement, blue: TOTO-3 nuclear stain. Coincidence of green and red signals produces a yellowish-white color. C: Control uninfected tissue section processed in parallel with the antibody combination used in A, B. Cell nuclei visualized with TOTO-3 nuclear stain. D, E: Detection of HSV-1 and HIV-1 antigens in close proximity at the ectocervical surface; green: rabbit polyclonal anti-HSV antibody, red: mouse monoclonal anti-HIV-1 p24 gag antibody with TSA enhancement, blue: TOTO-3 nuclear stain. Arrows indicate co-localization of HSV-1 and HIV-1 antigens in giant cell forms. F-G: Confocal fluorescence microscopy with normal premenopausal endocervical tissue that was cultured for 6 days prior to fixation and processing; green: rabbit polyclonal anti-cytokeratin antibody detecting epithelial cells, red: mouse monoclonal anti-CD4 antibody, with TSA enhancement, detecting a focus of CD4+ T cells located just below the epithelial surface, blue: TOTO-3 nuclear stain. Bars A = 25 µm; B = 20 µm; C = 50 µm; D = 50 µm; E = 20 µm; F = 50 µm; G = 25 µm.

Figure 4. Detection by in situ hybridization of HIV-1-infected cell aggregates at mucosal surfaces damaged by prior infection with HSV-2 and immunohistochemical detection of CD3 positive T cells.

In situ hybridization to detect HIV-1 RNA visualized by the accumulation of black silver grains; CD3 positive T cells visualized with DAB (brown stain); HSV-2 infected cells visualized in the form of multinucleated giant cells. In this experiment, the anti-CD3 antibody detects either CD4+ or CD8+ T cells and both of these T cell populations can be found in proximity to the luminal surfaces in the FRT (see Figure 7 also). A: HSV-2+HIV-1-infected tissue. B: Control tissue infected only with HSV-2, showing minimal accumulation of silver grains. C, D: HSV-2+HIV-1-infected tissue. Inset in D shows a rare cell productively infected with HIV-1. Note that both the density and number of silver grains are higher for the productively infected cell than observed for the multinucleated giant cells. Bar = 25 µm.

Figure 5. Inflammation at the mucosal surface in human premenopausal cervix.

A: H&E staining of the squamocolumnar cervical transformation zone showing extensive inflammation and Nabothian cysts. Bar = 500 µm. Based on the evaluation scheme used for inflammation, this particular tissue was rated: Endocervix: Severe, multifocal, peri-epithelial inflammation. Ectocervix: Moderate, focal, stromal inflammation. In this section, neutrophils, lymphocytes and plasma cells are the primary constituents of the cellular infiltrate. B: Higher magnification image of the endocervical surface (area immediately below the arrowhead in panel A) showing the peri-epithelial location of the inflammatory cells. Bar = 100 µm.

Figure 6. Summary of inflammation observed in the ectocervix in proximity to the squamocolumnar cervical transformation zone and in the endocervix of normal premenopausal women.

Figure 7. Characterization of inflammatory cell populations in normal human premenopausal endocervix.

5 µm tissue sections were incubated with the primary monoclonal antibodies listed below. Positive signal (brown stain) was revealed with biotinylated secondary antibodies and streptavidin-peroxidase conjugates. Slides were counterstained lightly with hematoxylin. Images are representative of staining patterns observed with 15–20 independent tissue samples. A: CD3. B: CD4. C: CD45RO. D: Control Primary Antibody. A–D from Cervix#1; Bar = 100 µm. E: CD3. F: CD4. G: CD45RO. H: Control Primary Antibody. E–H from Cervix#2; Bar = 50 µm. I: CD3. J: CD4. I–J from Cervix#3; Bar = 50 µm. K: CD3. L: Claudin-4, detecting tight junctions between columnar epithelial cells. K–L from Cervix#4; Bar = 50 µm.