and Women'sHospital, Harvard Medical School, Boston,. Massachusetts. Mechanisms underlying human immunodeficiency virus type I (HIV-I) infection of the ...
American Journal of Pathology, Vol. 139, No. 1, July 1991 Copyright © American Association of Pathologits
Orchitis and Human Immunodeficiency Virus Type 1 Infected Cells in Reproductive Tissues from Men with the Acquired Immune Deficiency Syndrome Jeffrey Pudney and Deborah Anderson From Fearing Research Laboratory, Department of Obstetrics, Gynecology and Reproductive Biology, Brigham and Women's Hospital, Harvard Medical School, Boston,
Mechanisms underlying human immunodeficiency virus type I (HIV-I) infection of the male reproductive tract and the sexual transmission of HIV-1 through semen are poorly understood. To address these issues, the authors performed morphologic and immunocytochemical analyses of reproductive tissues obtained at autopsy from 43 male acquired immune deficiency syndrome (AIDS) patients. Monoclonal antibodies recognizing different subpopulations of white blood cells were used to detect leukocyte infiltration and map the location of potential lymphocytic/monocytic HIV-I host cells; and immunocytochemistry and in situ hybridization techniques were used to detect HIVV-I--infected cells in the testis, excurrent ducts, and prostate. Distinct
pathologic changes were observed in a majority of testes of AIDS patients that included azoospermia hyalinization of the boundary wall of seminiferous tubules, and lymphocytic infiltration of the interstitium. The reproductive excurrent ducts and prostate appeared morphologically normal except for the presence offocal accumulations of white blood cells in the connective tissue stroma In the testis many white blood cells were shown to be CD4 +, indicating the presence of abundant host cells (T-helper/inducer lymphocytes and macrophages) for HIV-1. Furthermore macrophages and cells of lymphocytic morphology were observed migrating across the boundary walls of hyalinized seminiferous tubules to enter the lumen. In 9 of the 23 cases tested for HIV-I protein expression by immunocytochemistry, HIV-I + cells of lymphocytic/monocytic morphology were found in the seminiferous tubules and interstitium of the testis, epididymal epithelium and connective
tissue of the epididymis and prostate. One patient with epididymal blockage had accumulations of HIV-1-antigen-positive cells of macrophage morphology in the distended lumen of the efferent duct& There was no evidence of active HIV-1 infection in germ cells or Sertoli cells of the seminiferous tubules or other epithelial cells lining the excurrent ducts or prostatic glands. (Am JPathol 1991, 139:149-160)
Although the acquired immune deficiency syndrome (AIDS) has been established as a sexually communicable disease, the mechanisms underlying the venereal transmission of the causative agent, the human immunodeficiency virus (HIV-1), are poorly understood. In the male, for HIV-1 to appear in semen, either the virus must be produced locally by cells of the reproductive tract, or free virus or infected cells (most likely CD4 + lymphocytes or macrophages) must migrate across the bloodtestis barrier or epithelial layers of the excurrent ducts and accessory organs. As yet there is little information concerning the presence of HIV-1 in reproductive organs, or where and how infected host cells or free viral particles penetrate the blood-testis barrier or epithelial cell barriers to gain entrance to the lumen of the reproductive tract and subsequently appear in semen. Furthermore it is not known whether HIV-1 is capable of infecting the seminiferous epithelium of the testes or any of the other various epithelial cell types that line the male reproductive tract. This is important because it has been reported that cerSupported by contract CSA-88-020 from the Contraceptive Research and Development (CONRAD) Program, under a cooperative agreement with the United States Agency of International Development (USAID), which in tum receives funds for AIDS research from an interagency agreement with the National Institute of Child Health and Human Development
(NICHD). The views expressed by the authors do not necessarily reflect the views of the USAID or CONRAD. Accepted for publication March 11, 1991. Address reprint requests to Dr. Jeffrey Pudney, Fearing Research Laboratory, Brigham and Women's Hospital, 250 Longwood Ave., Boston, MA 02115.
Pudney and Anderson
AJPJuly 1991, Vol. 1239, No. I
tain murine retroviruses infect and undergo replication in the epithelium of the mouse epididymis.1 It is also necessary to determine whether infection with HIV-1 directly or indirectly affects the morphologic integrity of reproductive tissues. Pathologic changes associated with HIV-1 infection in reproductive organs could influence not only the ability of virions or infected host cells to infiltrate these tissues, but also their rate of entry into semen. This is extremely relevant because recent studies indicate that a large variation occurs in levels of virus that can be detected in the semen of HIV1-seropositive donors.2-- Therefore it is essential to identify conditions that affect the presence and abundance of HIV-1 in reproductive tissues of AIDS patients. One contributing factor that may enhance the number of potential HIV-1 host cells in reproductive tissues is male genital tract inflammation (GTI). There is evidence that GTI causes a significant elevation in levels of known HIV-1 host cells, CD4 + lymphocytes and macrophages, in semen,5 and that GTI is significantly associated with the presence of HIV-1 in semen of seropositive men.3'4 It is possible therefore that GTI may be an important cofactor involved in the sexual transmission of HIV-1, particularly in high-risk groups such as male homosexuals, who are known to have a greater incidence of venereal infections than male heterosexuals.6 Several recent reports have documented pathologic changes in testes of AIDS patients.7 13 These abnormalities include azoospermia, interstitial tissue fibrosis, hyalinization of seminiferous tubule boundary walls, and lymphocytic infiltration. As yet there has been no extensive investigation of other regions of the reproductive tract of AIDS patients to determine if pathologic changes also occur in the excurrent duct system or accessory glands. To address these issues, we have carried out a morphologic analysis of various reproductive tissues obtained from AIDS patients at autopsy. Light microscopic analyses were performed to describe further pathologic lesions in the testis as well as to determine if pathologic changes were present in the excurrent duct system and accessory glands. Immunocytochemistry studies were performed to assess the abundance and types of white blood cells in these tissues, as normal residents of the reproductive tissues and in inflammatory infiltrates, and to detect cells infected with HIV-1.
Materials and Methods
Subjects Reproductive tissues were obtained at autopsy from 43 male AIDS patients whose ages ranged from 21 to 72
years (median, 38 years). Thirty-two of these individuals were homosexual, five were bisexual, one was an intravenous drug user, and two were hemophiliacs. In the remaining three cases no known risk factor for HIV-1 infection was available. Morphologically normal reproductive tissues from men not known to be HIV-1 seropositive or symptomatic for AIDS (presumably noninfected) were obtained at surgical orchiectomy for treatment of prostatic cancer (six) or at autopsy (two). The men's ages ranged from 19 to 61 years, and all testis showed evidence of active spermatogenesis with mature sperm. These tissues were run as control tissues in all studies.
Light Microscopy Reproductive tissues were obtained both retrospectively and prospectively. The retrospective study was carried out on 10% formaldehyde-fixed paraffin-embedded tissues (22 patients) that were provided by the Department of Pathology, Deaconess Hospital, Boston, Massachusetts. These tissues consisted predominantly of testes with occasionally a piece of the epididymis. For the prospective study, the testis with intact epididymis attached and slices of prostate were retrieved at the time of autopsy (21 patients). These tissues were obtained through the courtesy of the Pathology Departments at the Deaconess, Brigham and Women's, Beth Israel, and Boston City Hospitals (Boston, MA). The testis and epididymis were dissected to provide portions of the peripheral testis, rete testis, efferent ducts, initial segment of the epididymis, caput epididymis, corpus epididymis, cauda epididymis, and vas deferens. Tissues then were either fixed in 10% formaldehyde (24 hours at 4°C) and processed for low temperature (520 to 560C) embedding in paraffin wax, or in 2.5% glutaraldehyde in 0.1 mol/l (molar) phosphate buffer (24 hours at 40C) and processed for embedding in glycol methacrylate (JB4 Polysciences, Warrington, PA). Wax blocks were sectioned at 5 ,u and sections stained with hematoxylin and eosin. Methacrylate-embedded tissues were sectioned at 2 to 3 ,u and sections stained with toluidine blue. In six AIDS cases and eight control cases, pieces of unfixed testis were embedded in ornithine-carbamoyltransferase (OCT) compound (Tissue-Tek, Miles Laboratories Inc., Elkhart, IN) and rapidly frozen in hexane at - 700C. Also several epididymides from AIDS patients were dissected into the caput, corpus, and cauda regions, fixed in 4% paraformaldehyde, cryoprotected in 15% sucrose at 40C, embedded in OCT compound, and frozen in hexane at - 700C. All tissue blocks were stored at - 700C. To obtain a representative analysis of the reproductive organs, several pieces of tissue were processed from each part of
Orchitis in Reproductive Tissues of Men with AIDS
AJP July 1991, Vol. 139, No. 1
the genital tract. Several sections were then cut from each of these blocks for histologic examination.
Immunocytochemistry White Blood Cell Markers Wax sections were deparaffinized in xylene (four times, 10 minutes each) and rehydrated to 95% alcohol. Optimal titers for each primary antibody were determined in a pilot study using formalin-fixed paraffin-embedded lymph node as a target tissue; antibodies were diluted in 0.02 mol/l TRIS buffer, pH 8.2, containing 1% bovine serum albumin and 0.02 mol/l sodium azide (Sigma Chemicals, St. Louis, MO). After a short rinse (2 minutes) in 0.05 mol/l TRIS buffer, ph 7.6, sections were incubated with the following primary monoclonal antibodies (MAb) that recognize antigens in wax sections: Dako leukocyte common (LC) MAb (Dako, Santa Barbara, CA) diluted 1:40, which recognizes an antigen that is expressed on all white blood cells (but especially reacts with lymphocytes in wax sections), and Dako L26 diluted 1:40, which reacts with B lymphocytes. Frozen sections of testes and epididymides were cut on a cryostat at a thickness of 4 to 5 ,u, and allowed to air dry on gelatin-coated slides. Sections of the testis were fixed in acetone for 10 minutes at room temperature. The sections then were reacted with a panel of MAbs to identify specific subpopulations of white blood cells: anti-macrophage (Dako-macrophage), antiCD8, found primarily on T-suppressor/cytotoxic lymphocytes (Dako anti-T8), anti-Cd4, found primarily on Thelper/inducer lymphocytes and macrophages (Dako OKT4, and anti-3a + b, Becton-Dickinson, San Jose, CA), and anti-CD 2,3 a pan T-lymphocyte marker (antiLeu 4a + 5b, Becton-Dickinson). All primary antibodies were run at a dilution of 1:40, and were incubated with the sections for 30 minutes at room temperature (RT). Wax and frozen sections were then rinsed in TRIS buffer and the antibodies visualized by means of a mouse monoclonal alkaline phosphatase/anti-alkaline phosphatase (AP/ AAP) detection system (Dako) that stains positive cells red. Sections were finally counterstained with aqueous hematoxylin. As negative controls, tissues were either processed in the absence of the primary antibody, or with a comparable concentration of nonreactive antibody of the same isotype. HIV Marker
Cells infected with HIV were identified in wax sections of reproductive tissues using MAbs and either an immunoperoxidase or an AP/AAP detection system. For the immunoperoxidase test, sections were deparaffinized,
rehydrated to 95% alcohol, and then placed in 80% methanol containing 3% H202 for 20 minutes to quench endogenous peroxides. They were then rinsed in phosphate-buffered saline (PBS) and digested in prewarmed 0.01 N HCI containing 0.4% pepsin (P-7000; Sigma Chemical Co., St. Louis, MO) for 12 minutes at 37°C to unmask immunoreactive sites. After a wash in PBS, the sections were exposed to a primary mouse MAb directed against the HIV-1 core protein p24 (DuPont, Wilmington, DE; diluted 1:10) at 40C for approximately 18 hours. The sections were then allowed to reach RT, washed with PBS, and incubated with biotinylated horse anti-mouse gamma G immunoglobulin (Vector, Burlingame, CA; diluted 1:500) for 20 minutes. After rinsing in PBS, sections were reacted with a streptavidin horseradish peroxidase conjugate (Vector; diluted 1:1,000) for 20 minutes. The primary antibody was then visualized by incubating the sections in a 0.4% solution of 3-amino-9-ethylcarbozole (Sigma) in 0.1 mol/l acetate buffer pH 5.2 containing 3% H202 for 20 minutes. For the AP/AAP test, sections were rehydrated and digested with 0.4% pepsin, after which they were incubated with a cocktail of antibodies to the HIV-1 antigens, p24, gp120, and gp4l (DuPont; diluted 1:20) at 4°C for 18 hours. These antibodies then were visualized using an AP/AAP detection kit (Dako). Finally all sections were counterstained with aqueous hematoxylin. Negative controls included reproductive tissues from non-HIV-1-infected men, and omission of the primary antibody; positive controls consisted of lymph nodes from HIV-1-infected men and a Jurkat cell line infected with HIV-1 Illb that had been fixed in 10% formaldehyde and then processed for low-temperature embedding in wax. In Situ Hybridization In situ hybridization was carried out using an Insite kit (supplied by Applied bioTechnology, Inc., Worcester, MA) especially designed for identifying HIV-1 in formaldehyde-fixed, wax-embedded tissues. The procedure was carried out according to the manufacturer's instructions. In brief, tissue sections were dewaxed in xylene, rehydrated to 70% alcohol, sequentially washed with PBS and 0.1 mol/l glycine in PBS and then immersed in 0.2 N HCI for 10 minutes at RT. This was followed by digestion in Proteinase K for 10 minutes at RT. Sections then were washed in glycine/PBS, fixed in formaldehyde for 10 minutes, washed with PBS, and finally allowed to hybridize with a biotinylated probe specific for HIV-1 mRNA for 18 hours at 37°C. After washing in various concentrations of saline sodium citrate (SSC) buffer (4x and 2x at 370C, and 0.1 x and 4x at RT), the sections were incubated with the enzyme solution (streptavidin alkaline phosphatase) for 30 minutes at RT. The sections then were
Pudney and Anderson
AJP July 1991, Vol. 139, No. 1
placed in a wash solution, followed by two rinses in 4x SSC and the probe visualized by reacting sections with a nitroblue tetrazolium substrate for 30 minutes at 370C. This procedure results in HIV-1-positive cells appearing purple-blue to black. Controls included those supplied with the kit (wax-embedded peripheral lymphocytes from an HIV-1-seropositive donor [positive control] and an HIV-1-seronegative donor [negative control]), as well as formaldehyde-fixed wax-embedded Jurkat cells infected with HIV-1 Illb.
Morphologic Assessment by Light Microscopy Testis
Of the 43 testes examined from AIDS patients, 29 were azoospermic, with the seminiferous epithelium reduced to either a layer of pachytene spermatocytes and a basal layer of leptotene spermatocytes (Figure 1 a), or virtually Sertoli cells only (Figure 1 b). Many of the germ cells present in the seminiferous tubules were degenerating, as evidenced by the presence of pyknotic nuclei. Twelve of the testes displayed impaired spermatogenesis but contained seminiferous tubules with some residual spermatids. Only two of the testes contained seminiferous tubules that appeared to be undergoing active spermatogenesis; however even in these testes the lumen of many seminiferous tubules was filled with sloughed immature germ cells. In most testes examined, areas of hyalinized seminiferous tubules were observed. This ranged from moderate thickening of the boundary wall with the seminiferous epithelium intact (Figure 1 c) to overt hyalinization, where there was massive thickening of the boundary wall and complete loss of the seminiferous epithelium (Figure 1 d). Distinct focal accumulations of small mononuclear cells could be detected in the interstitial tissues (Figure 1 e). At the light microscope level, these cells principally resembled lymphocytes. The lymphocytic accumulations were scattered throughout the interstitium of the testis but were most prevalent surrounding large blood vessels and areas of hyalinized tubules. Every control testis examined contained seminiferous tubules with intact germinal epithelia. All tubules displayed active spermatogenesis with a mosaic pattern of germ cell development typical of the normal human testis (Figure 3f). No focal accumulations of lymphoid cells were observed in interstitial areas. A few hyalinized seminiferous tubules were observed in only one of the testes examined.
Excurrent Ducts and Prostate
At the light microscopic level, the epithelium lining the excurrent ducts and prostate specimens from AIDS patients appeared intact; no obvious pathologic changes could be detected in these tissues (Figure 3a-c). In the epididymis, particularly in regions proximal to the testis, round cells distinctly different from epithelial cells could be seen within the apical regions of the epithelium. These cells were detected at many levels within the epididymal epithelium, suggesting that they were migrating across this tissue to the luminal aspect of the epithelium. Some of these cells seemed to breach the epithelium and enter the lumen of the epididymis (Figure if). In all cases examined, a distinctive feature was the presence of blood vessels located within the boundary wall of the excurrent ducts or prostatic glands. In many instances these blood vessels were clearly opposed to the basal lamina of the epithelium of these tissues and occasionally could even be found located within the epithelium (Figure 1 g). Excurrent ducts from control (non-AIDS) men also appeared histologically normal and the tubules were distended with spermatozoa (Figure 3g). The intraepithelial cells were not as prevalent in these cases.
Immunocytochemistry: White Blood Cell Markers Testis Only tissues that appeared morphologically well preserved were used for immunocytochemical studies. Paraffin wax sections from 23 different testes from AIDS patients were reacted with the Dako LC MAb, which recognizes white blood cells, especially lymphocytes, in wax sections. Abundant LC + cells were detected in the interstitial tissue in a majority (74%) of cases (Figure 2a). Commonly they occurred as distinct focal accumulations (Figure 2a) and presumably corresponded to the groups of cells previously observed by light microscopy. LC + cells were abundant surrounding blood vessels and hyalinized seminiferous tubules. We also observed apparent invasion of seminiferous tubules by LC+ cells. In many instances, LC+ cells were detected within the boundary wall and reduced epithelium of seminiferous tubules (Figure 2b). Leukocyte common positive cells were even more prevalent in the thickened boundary wall and lumen of completely hyalinized seminiferous tubules (Figure 2c). Only a few cells positive for the B lymphocyte marker could be detected in the interstitial tissue of the testes. These cells were usually found within the accumulations of cells previously found to be positive for the LC marker. Omission of the primary antibody as a nega-
Orchitis in Reproductive Tissues of Men with AIDS
Vol. 139, No. 1
Figure 1. Light microscopy, of reproductive tissues from AIDS patients. a: Azoospermic testis where the seminiferous epithelium has been reduced to a layer of pachytene spermatocytes (arrouheads). AMethacrylate section. Toluidine blue (x 120). b: Azoospermic testis with seminiferous epithelium composed mostl/ of Sertoli cells. Methacrylate section. Toluidine blue (x 560). C: Incipient hyalinization. Boundary wall of seminiferous tubules is moderately thickened and seminiferotus epithelium is intact. Mfethacrylate section. Toluidine blue (X80). d: Complete hyalinization. Boundary uwall of seminiferous tubules is massively thickened and seminiferous epithelium is no longer present. Methacrylate section. Toluidine blue (x270). e: Focal accumulation of cells, mostly resembling lymphocy'tes, present in the interstitium. Methacrtlate section. Toluidine blue (x200). f: Initial segment of epididymis showing a cell that appears to be leaving the epithelium to enter the lumen. Wax section.(H&E X630). g: Caput region of the epididqjmis demonstrating the presence of a blood vessel (arrowheads) actually located within the epithelium. Wax section. (H&E X 480).
tive control demonstrated no background or nonspecific
were the most abundant white blood cells present. They
staining (Figure 2d). Immunohistologic examination of frozen sections of testes from six AIDS patients showed that macrophages
occurred throughout the interstitial tissue and completely surrounded the seminiferous tubules (Figure 2e). Also cells staining positive for the macrophage marker were
Pudney and Anderson
AJPJuly 1991, Vol. 139, No. 1
0. s, 4, ..:2.z
b ?I 0. i
* } }._508=,§!s,,.
Orchitis in Reproductive Tissues of Men with AIDS
155 AJPJuly 1991, Vol. 139, No. 1
detected within the thickened boundary wall and lumen of hyalinized seminiferous tubules. Many large pale staining cells and small intensely staining cells were observed in the interstitium after reaction with either of the anti-CD4 antibodies. Similarly many small intensely staining cells were observed when the pan T-lymphocyte MAb was used (Figure 2f). In all testes examined from AIDS patients only a few CD8+ cells were observed either in small accumulations or dispersed throughout the interstitium (Figure 2g). In wax sections of testes from non-HIV-1-infected men, only a few scattered LC + cells and no B cells were detected in the interstitial tissue (Figure 3f). In frozen sections, many macrophages but few T lymphocytes were observed in the interstitial tissues of the testis, and no macrophages or lymphocytes were observed within the boundary walls, seminiferous epithelium, or lumen of the seminiferous tubules. Excurrent Ducts and Prostate Reproductive excurrent ducts and prostates from 19 HIV-1-positive patients were examined for the presence of white blood cells. LC + cells were present in the efferent ducts, segments of the epididymis, and the vas deferens as well as the prostate of all AIDS subjects, and were abundant in 53% of cases (Figure 3a-c). These cells were most numerous in the connective tissue stroma surrounding the excurrent ducts and often occurred as large accumulations close to the tubules of these organs. Another consistent observation was the presence of accumulations of LC + cells surrounding blood vessels. LC + cells also frequently could be detected within the epithelium lining the excurrent ducts and prostatic glands (Figure 3b-d). Most often these cells occurred in the basal region of the epithelium. Sometimes, however, LC + cells could be observed deep in the epithelial layer (Figure 3d) and lumen of the epididymis or prostate. Incubation of frozen sections of epididymal regions with the panel of MAbs for identifying subpopulations of white blood cells showed that the most abundant cell type present was the macrophage. These cells were found within the epithelium and connective tissue (Figure 3e). Lymphocytes were identified in the connective tissue and epithelium of the epididymis after incubation with the
pan-T-cell marker. Few of these lymphocytes were positive for the CD8 marker or B-cell marker. Control epididymides also contained a few LC + cells, primarily macrophages and a few CD4 + and CD8 + lymphocytes, in the tissue stroma of the epididymis (Figure 3g). Only sometimes were these cells seen in the epithelial layer and lumen.
Immunocytochemistry: HIV Markers Of 23 AIDS patients analyzed, only nine had cells in reproductive tissues that reacted with antibodies to the HIV1 antigens p24, gpl 20, and gp4l. In these patients, P24 + cells were occasionally found in the interstitial tissue of the testis (Figure 4a), and connective tissue stroma of the prostate (Figure 4b). When the cocktail of HIV-1 antibodies was used, however, several positive cells were detected within a few seminiferous tubules (Figure 4c) and epithelium of the epididymis (Figure 4d). These cells were all of lymphoid morphology. In one patient with evidence of blockage in the epididymis, numerous cells staining positive for HIV-1 antigens could be detected in the distended lumen of the efferent ducts that was filled with spermatozoa (Figure 4e). These large HIV1-positive cells resembled macrophages, and engulfed sperm heads could often be observed in their cytoplasm. There was no evidence for active infection of Sertoli cells, germ cells, or any of the epithelial cells lining the excurrent ducts and prostate. Negative control tissue sections from noninfected men and sections from AIDS patients processed in the absence of the primary antibody or with isotypic control antibody consistently showed negligible background or nonspecific staining. Positive controls consisting of a Jurkat cell line infected with HIV-1 Illb demonstrated good specific staining when reacted with the antibody to p24 (Figure 4f). In Situ Hybridization Six cases that had HIV+ cells by immunocytochemistry were investigated for the presence of HIV-1 using in situ hybridization. Of these, only one contained cells that were positive. This was the patient with epididymal obstruction and the most numerous HIV-1 + cells by immu-
Figu re 2. Immunocytochemistry of testesfrom AIDS patients. a: Numerous LC + cells occurred in the interstitium ofthe testis and were often present as large accumulations (x 400). b: 1Hyalinized seminiferous tubules with reduced seminiferous epithelium. LC+ cells are located within the thickened boundary wall and also within the seminiferous epithelium (x 790). c: Completely byalinized seminiferous tubule with numerous LC + cells within the boundary wall and in the lumen (arrowheads) (x 780). d: Control testis tissue where the primary antibody (LC) was omitted. Note absence of background or nonspecific staining (x 180). e: Frozen section of testis reacted with the Dako macrophage monoclonal antibodv. Abundant positive cells are present in the interstitium surrounding the seminiferous tubules (x 150). f: Frozen section of testis incubated with the Leu 4a + 5b (Pan T cell) monoclonal antibody demonstrating many positive cells in the interstitium (x300). 9: Frozen section of testis incubated with the Leu 2 (anti-CD8) monoclonal antibody showingfew T lymphocytes (of the T8 phenotype) in the interstitium (x300).
Pudney and Anderson
AJP July 1991, Vol. 139, No. 1
Orchitis in Reproductive Tissues of Men with AIDS 157 AJPJuly 1991, Vol. 139, No. 1
nocytochemistry. Human immunodeficiency viruspositive cells were detected in the same region of the efferent ducts previously shown to contain cells that reacted positively to the cocktail of HIV-1 antibodies (Figure 4g). The negative controls had no background or nonspecific staining, whereas both positive controls stained intensely for the presence of HIV-1 or HIV-1 Illb.
Discussion The results of this study demonstrate that white blood cells commonly infiltrate the testes, excurrent ducts, and prostates of AIDS patients. The focal accumulations of white blood cells were initially detected in paraffin wax sections of these reproductive organs using an MAb (LC) recognizing an antigen expressed by all white blood cells, but especially on lymphocytes in wax sections. Morphologically these cells resembled lymphocytes. To identify further the white blood cells, frozen sections of the testis and epididymis were reacted with a panel of MAbs directed against specific antigens present on different populations of white blood cells. The most abundant white blood cells detected in the testis were macrophages, which were dispersed throughout the interstitial tissue and completely surrounded the seminiferous tubules. Furthermore macrophages were observed also in the thickened boundary wall and lumen of hyalinized seminiferous tubules. Because macrophages express a small number of CD4 receptors, they also reacted weakly with the anti-CD4 antibodies. Many smaller, more intensely stained cells, however, were also detected with anti-CD4 antibodies, and these cells were also positive for the pan-T-lymphocyte marker. Only a minority of T cells in the testis were CD8+, and very few B lymphocytes were observed. These findings indicate the existence of numerous CD4 + T lymphocytes in the testes of HIV-1-positive patients. Similar results were also found in the epididymis, where the most prevalent white blood cells present were macrophages. These macrophages, as well as lymphocytes, were observed in the connective tissue and epithelium of the epididymis. Few of the lymphocytes were positive for the T8- or B-cell markers, in-
dicating that most of these cells helper/inducer phenotype.
of the CD4 T-
These observations are distinctly different from the ocand organization of immunocompetent cells in the normal male reproductive tract.14 It has been previously reported that numerous macrophages occur in the interstitium of the healthy human peripheral testis, whereas lymphocytes of any phenotype are virtually absent, but can be found in the rete testis.1"18 Analysis of control testes and epididymides from noninfected men in the present study confirmed these observations. Therefore the accumulations of white blood cells, many of which were CD4 + T lymphocytes, in a majority of testes and other reproductive tissues examined from AIDS patients is abnormal. It is interesting to note, however, that in a study of infertility patients, CD4 + T lymphocytes were found in the testes of men with unilateral testicular obstruction or vasectomy.15 The morphologic appearance of the testes in this study, ie, azoospermia, hyalinization of the boundary wall of the seminiferous tubules, and lymphocytic infiltration of the interstitial tissues indicates that testicular orchitis may be a common condition of AIDS patients. The presence of lymphocytic infiltrates in the excurrent ducts also suggests that subclinical inflammation in these regions may also be associated with HIV-1 infection. Many systemic viral infections are capable of causing orchitis.19 It is not yet known whether the orchitis in HIV-1-positive patients is a nonspecific sequelae of viremia, is attributable to a direct effect of HIV-1 on reproductive tissues, or attributable to other factors. In some experimental animal models, neonatal thymectomy produces a T-cell immunodeficiency in adults that is associated with orchitis and epididymitis, presumably because immunoregulatory mechanisms that normally prevent such autoimmune manifestations are disrupted.20 Therefore orchitis in AIDS patients may be due to their immunodeficient status. Regardless of the cause, the lymphocytic infiltration of reproductive tissues of AIDS patients that includes numerous CD4+ natural host cells for HIV-1 could clearly enhance the levels of HIV-1 in the reproductive tract and appearance of HIV-1-infected cells in semen. Passage of infected cells across reproductive epithelia would be currence
Figure 3. Immunocytochemistry of excurrent ducts and prostate from AIDS patients (a-e) and of testis and epididymis from control, non-AIDS patients (f, g). a: Efferent ducts. Abundant LC+ cells were detected in the connective tissue stroma surrounding the efferent ducts. The epithelium lining the ducts is intact and shows no obvious pathologic changes (X520). b: Initial segment of epididymis, demonstrating the presence of numerous LC + cells in the connective tissue stroma. Note the LC + cell present within the basal region of the epididymal epithelium (arrowhead). The epididvmal epithelium is also intact and no gross structural abnormalities could be detected (x 600). c: Prostate showing (a) many LC + cells in the connective tissue stroma, and (b) intraepithelial LC + cells (arrowheads). The morphologic appearance of the prostate was normal and no pathologic changes were observed at the light microscope level (x375). d: Caput epididymis. LC + cells were often detected in the basal region of the epithelium and occasionally deep within the epithelium (x590). e: Frozen section of caput
epididymis incubated u'ith the Dako macrophage monoclonal antibodv showing numerous large positive cells present in the epithelium (X215). f: Testis from normal HIV-1-seronegative patient reacted with the LC antibody. Few LC+ cells could be detected in the interstitial tissue (x 135). g: Epididymis firom normal HIV-1-seronegative patient reacted with the LC antibody. Few LC + cells are present in the interstitial stroma or epithelium of this section through the corpus epididymis (x 130).
Pudney and Anderson
AJPJuly 1991, Vol. 139, No. 1
Figure 4. Detection of HIV-I in reproductive tissues ofAIDS patients bl immunocytochemistrv (a-f) and in situ hybridization (g). a: A single P24 + cell detected in the interstitial tissue of the testis. Immunoperoxidase (x 710). b: A group of P24 + cells observed in the connective tissue of the prostate. Immunoperoxidase (5x 700). C: Cell staining positive with the cocktail of anti-HNV-I antibodies present in the seminiferous epithelium. AP/AAP (X 680). d: Cell staining positive with the cocktail of anti-HIV-1 antibodies present in the epithelium of the caput epididymis. APIAAP (x 760). e: Distended efferent ductsfilled with spermatozoa, due to possible epididymal blockage, contained many large cells that reacted positive with the cocktail of anti-HV- 1 antibodies. APIAAP (x 480). f: Positive cell controlfor p24 monoclonal antibod . Jurkat cells infected with HNV-IIIb stained intensely when exposed to the P24 antibody. APIAAP (x220). g: HV-1-positive cells were also observed in the distended efferent ducts of the patient with epididy!mal obstruction by in situ hybridization (x5600).
facilitated by the close apposition of blood vessels to the basal lamina of these tissues that was observed in most reproductive organs examined in this study. Of particular importance was the observation of both macrophages
and T lymphocytes within the boundary walls and lumina of seminiferous tubules, particularly those that had become hyalinized. This suggests that the blood-testis barrier is compromised or destroyed in AIDS patients, en-
Orchitis in Reproductive Tissues of Men with AIDS
AJP July 1991, Vol. 139, No. 1
abling these cells to invade the seminiferous tubules. Therefore, as AIDS patients become progressively azoospermic with hyalinization of seminiferous tubules, there develops a potential for HIV-infected host cells to enter the lumina of these tubules to appear in the semen. Several seminiferous tubules were found, in fact, to contain cells that reacted positively for HIV-1 antigens. This finding relates to other studies showing that the prevalence of HIV-1 in semen increases as the disease stage advances.34 Intraepithelial lymphocytes and macrophages were often detected in the efferent ducts, epididymis, and prostate. Only a few of these cells appeared to be migrating through the epithelium, however, and only an occasional lymphocyte could be seen in the lumina of these organs. As yet the identity of the round cells observed by light microscopy in the epithelium of the proximal region of the epididymis is unknown. The results of this study confirm earlier reports that distinct pathologic lesions are commonly present in testes of AIDS patients,7-13 but extends these studies by showing that other regions of the reproductive tract, the excurrent duct system and prostate, appear morphologically normal. The extent of testicular atrophy differed within the group of AIDS patients examined. Although approximately 68% displayed complete azoospermia, nearly 28% possessed testes that, although not undergoing active spermatogenesis, did contain seminiferous tubules with some residual spermatids present. Only 4% of the testes analyzed exhibited any normal spermatogenic activity. This variation could reflect a progressive loss of testicular function in AIDS patients. The precise cause of the decline in spermatogenesis in AIDS patients remains to be determined. It has been suggested that atrophy of the testis is related to chronic disease, with attendant fever and malnutrition, all of which are factors found in terminally ill AIDS patients. In addition, the effects of therapeutic drugs received by AIDS patients on testis structure and function are unknown. Furthermore, reproductive tissues may be infected with other pathogens in immunodeficient AIDS patients. Given these factors, it is difficult to dissect out what direct effect, if any, infection with HIV-1 has on the male reproductive tract.
None of the patients in the present study were receiving treatments known to adversely affect spermatogenesis. Also regardless of the age at death, or duration of AIDS from initial diagnosis to death (which was as short as 3 months but was usually between 8 and 11 months), nearly 96% of these individuals displayed testicular atrophy. Furthermore few of these AIDS patients were classified as chachetic. When related to other studies where patients were also screened for factors deleterious to the testis,12,13 it appears that the only common determinants linking all individuals in these separate studies are infection with HIV-1 and immunodeficiency.
It is also intriguing that the most striking AIDSassociated pathologic condition within the reproductive tract is disruption of spermatogenesis. In the present study, the excurrent ducts and prostate appeared normal. Because these tissues are androgen dependent, it can be assumed that Leydig cell function in AIDS patients is not grossly disturbed by HIV-1 infection. As yet there is no information on whether the pituitary is adversely affected by HIV-1 infection, and there are few studies on circulating levels of gonadotropins and testosterone in AIDS patients. One study, limited to HIV1-infected male subjects asymptomatic for AIDS, reported that both testosterone and luteinizing hormone levels were elevated in these individuals, whereas folliclestimulating hormone levels were unchanged when compared with controls.21 As yet, however, peripheral levels of these hormones in symptomatic AIDS patients have not been determined. Therefore an endocrine disturbance could be responsible for azoospermia in AIDS patients. It has been reported that spermatogenesis can be impaired in humans after infection with different viruses,19 and infection of the male genital tract with cytomegalovirus has been associated with a reduction in sperm count.22 The possibility exists, therefore, that testicular atrophy in AIDS patients may result from a direct effect of HIV-1 on the testis. Viruses, however, are too large to passively infiltrate the blood-testis barrier, and active transport of viral particles through Sertoli cells or by other mechanisms into the germ cell compartment have not yet been described. Alternatively disruption of spermatogenesis in AIDS patients could arise indirectly from infection with HIV-1 due to orchitis as described above. Very few HIV-1-infected cells were observed in reproductive tissues of AIDS patients. Some cells positive for HIV-1 antigens were detected by immunocytochemistry in the seminiferous tubules and interstitial tissue of the testis, connective tissue stroma of the epididymis and prostate, and epithelium of the epididymis. These cells were all lymphoid in appearance and corresponded to areas where CD4 + lymphocytes and macrophages were observed. The in situ hybridization technique for detection of HIV-1-infected cells was not as sensitive as immunocytochemistry in this study, possibly because of degradation of messenger RNA (mRNA) in autopsy tissue. In one patient with a previous history of gonorrhea, there was an apparent epididymal obstruction that resulted in accumulation of spermatozoa and immature germ cells in the highly distended lumen of the efferent ducts. Also present were numerous cells resembling macrophages that were undergoing phagocytosis of germ cells. Many of the macrophages were positive for HIV-1 by both immunocytochemical and in situ hybridization techniques. These data suggest that both the testis and excurrent ducts can act as sites of entry of HIV-
Pudney and Anderson
AJP July 1991, Vol. 139, No. 1
1-infected host cells into the lumen of the reproductive tract. There was no evidence for production of HIV-1 by Sertoli cells, germ cells, or other epithelial cells lining the excurrent ducts or prostate. The immunocytochemical and in situ hybridization techniques used, however, can only detect HIV-1 proteins and mRNA, respectively. Therefore the host cell must be activated and producing viral progeny before it can be identified by these methods. Latently infected cells containing HIV-1 in the proviral form were not detected in this study. In future studies, DNA hybridization techniques such as in situ hybridization at the electron microscopic level, or polymerase chain reaction procedures should be applied to selected cell types isolated from reproductive tissues to determine if proviral HIV-1 is present in reproductive cells. In addition, the present study only investigated the testis, excurrent ducts, and prostate; further studies should be performed to characterize potential HIV-1-infected cells in other male reproductive tissues such as the seminal vesicles, urethra, and urethral glands. Clearly more information is needed before the biologic relationship between HIV-1 infection and the male reproductive tract can be completely understood. The successful management and control of AIDS by the development of drugs or vaccines will partly depend on understanding exactly how the virus is sexually communicated, where it or infected host cells reside in the male reproductive tract, and its effects on male reproductive tissues.
9. 10. 11.
12. 13. 14.
The authors thank Chris Simpson (Laboratory for Human Reproduction and Reproductive Biology) for technical assistance, Dr. C.J. O'Hara, Don Delutis, and the staff of the Pathology Department, Deaconess Hospital, Boston, Massachusetts, for advice and for providing access to pathologic material, and Julie Kelleher (Fearing Laboratory) for the preparation of this manuscript.
References 1. Kiessling AA, Crowell RC, Connell RS: Sperm associated retrovirus in the mouse epididymis. Proc Natl Acad Sci USA
1987, 84:8667-8671 2. Van Voorhis BJ, Martinez A, Mayer K, Anderson DJ: Detection of human immunodeficiency virus type 1 in semen from seropositive men using culture and polymerase chain reaction deoxyribonucleic acid amplification techniques. Fertil Steril 1991, 55:588-594 3. Anderson DJ, Wolff H, Pudney J, Wenhao Z, Martinez A, Mayer K: Presence of HIV in Semen. Heterosexual Transmission of AIDS. In Alexander NJ, Gabelnick HL, SpielerJM, eds. New York, NY, Wiley-Liss, 1990, pp 167-180
4. Anderson DJ, Politch JA, Martinez A, Wolff H, Pudney J, Van
Voorhis B, Meyer K, Saltzman S, Seage G, Horsburgh CR: Prevalence and temporal variation of HIV-1 in semen. 6th International Conference on AIDS. San Francisco, CA, 1990. Wolff H, Anderson DJ: Male genital tract inflammation is associated with increased numbers of potential human immunodeficiency virus host cells in semen. Andrologie 1988, 20:404-410 Judson FN, Penley KA, Robinson ME, Smith JK: Comparative prevalence rates of sexually transmitted diseases in heterosexual and homosexual men. Am J Epidemiol 1980, 112:836-843 Reichert CM, O'Leary TJ, Levens DL, Simrell CR, Macher AM: Autopsy pathology in the acquired immune deficiency syndrome. Am J Pathol 1983,112:357-382 Welch K, Finkbeiner W, Alpers CE, Blumenfield W, Davis RL, Smuckler EA, Beckstead JH: Autopsy findings in the acquired immune deficiency syndrome. JAMA 1984, 252:1152-1159 Mobley K, Rotterdam HZ, Lerner CW, Tapper ML: Autopsy findings in the acquired immune deficiency syndrome. Pathol Annu 1985, 20:45-65 Niedt GW, Schinella RA: Acquired immunodeficiency syndrome: Clinicopathologic study of 56 autopsies. Arch Pathol Lab Med 1985,109:727-734 Chabon AB, Stenger RJ, Grabstaldt H: Histopathology of testis in acquired immune deficiency syndrome. Urology 1987, 29:658-4663 Rogers C, Klatt EC: Pathology of the testis in acquired immunodeficiency system. Histopathology 1988, 12:659-665 DePaepe ME, Waxman M: Testicular atrophy in AIDS. A study of 57 autopsy cases. Hum Pathol 1989, 20:210-214 Pudney J, Anderson DJ: Organization of immunocompetent cells and their function in the male reproductive tract. In Johnson PM, Griffin PD, eds. WHO Symposium on Immunology of the Reproductive Tissues. (In press) El-Demiry MI, Hargreave TB, Busuttil A, Elton R, James K, Chisholm GD: Immunocompetent cells in human testis in health and disease. Fertil Steril 1987, 48:470-479 Pollanen P, Niemi M: Immunohistochemical identification of macrophages, lymphoid cells and HLA antigens in the human testis. Int J Androl 1987; 10:37-42 El-Demiry Ml, James K: Lymphocyte subsets and macrophages in the male genital tract in health and disease. Eur Urol 1988,14:226-235 El-Demiry Ml, Hargreave TB, Busuttil A, James K, Ritchie AWS, Chisolm GD: Lymphocyte subpopulations in the male genital tract. Br J Urol 1985, 57:769-774 Mikuz G, Damjanov I: Inflammation of the testis, epididymis, peritesticular membranes, and scrotum. Pathol Annu 1982, 17:101-128 Taguchi 0, Nishizuka V: Experimental autoimmune orchitis after neonatal thymectomy in the mouse. Clin Exp Immunol 1981, 46:425-434 Merenich JA, Asp AA, McDermott MT, Kidd GS: Gonadal function in men with asymptomatic human immunodeficiency virus (HIV) infection. Presented at the 71 st Endocrine Society Meeting, Seattle, WA, Abst. #581. 1989 Lang DJ, Kummer JF, Hartley DP: Cytomegalovirus in semen. Persistence and demonstration in extracellular fluids. N Engl J Med 1974, 291:121-123