Introduction

Mpox (formerly monkeypox) is a zoonotic disease caused by the monkeypox virus (MPXV), member of the Poxviridae family, which includes other viruses of significant public health importance including smallpox and vaccinia viruses1. In people, infection causes fever, pustular rashes, and lymphadenopathy, and less commonly pneumonia, encephalitis, ocular lesions, and death1. The virus spreads via respiratory droplets, direct contact with mucocutaneous lesions, vertically during pregnancy, and likely via sexual transmission given high viral loads in seminal fluid and semen1,2. While a majority of cases in the 2022 outbreak were reported in MSM, women are equally susceptible to the virus1,3. A number of cases in women were identified during pregnancy4, and asymptomatically, therefore lesions in women are less well described. As of August 14, 2024, mpox has been declared a public health emergency of international concern (PHEIC) by the World Health Organization due the emergence of new mpox clades that are spreading in the Democratic Republic of the Congo. Here we present data on viral replication in the reproductive tracts of rhesus macaques experimentally challenged with mpox and report a case of mpox replication associated with endometriosis in a female rhesus macaque.

Macaques have similar endometrial physiology to humans5 and frequently develop spontaneous endometriosis6,7 with comparable health and fertility sequelae8. In both women and macaques, endometriosis is characterized by proliferation of endometrial tissue outside the uterus, causing pain, peritoneal adhesions, and even complete infertility9. Endometriosis is estimated to affect up to 10% of reproductive-aged women, but it is likely underdiagnosed due to symptom heterogeneity and requirement for surgical visualization of lesions for definitive diagnosis9. Given the possibility of sexual transmission of mpox via semen, the uterus could be a prime target for infection.

Results

Mpox was identified in the reproductive tract of a female macaque with endometriosis

We previously reported that primary infection with mpox provided robust protection against mpox rechallenge10. As part of this rechallenge study, two female rhesus macaques and one male macaque were challenged intravenously with mpox as previously naïve controls, euthanized 10 days following challenge, and necropsied for tissue evaluation. As previously described, all naïve, mpox-infected macaques developed classic vesicular epidermal lesions throughout the body10. Lesions were visible by day 7 post-challenge and increased at day 9 (Table 1). At necropsy, gross lesions in the reproductive tract of a female macaque (Animal 1) were reported as “chocolate cysts” that expanded the uterus with proliferative tissue obscuring the oviducts and ovaries, consistent with endometriosis. No reproductive lesions were noted in the other female macaque (Animal 2) or the male macaque (Animal 3). Tissues from the reproductive tracts from mpox infected animals were screened for MPXV by immunohistochemistry (IHC) and the only animal with viral replication in reproductive tissues was the female macaque with endometriosis (Animal 1, Table 2).

Table 1 Pox lesion counts and distribution post-challenge.
Full size table
Table 2 Mpox IHC in reproductive tissues as compared to skin and tonsil of rhesus macaques.
Full size table

Mpox replication targets vimentin-positive stroma and poorly differentiated glandular epithelium in endometriotic lesions

Endometriosis in Animal 1 was best appreciated in evaluation of the ovarian and oviduct tissue (Fig. 1, Suppl. Fig. 1). As compared to normal oviduct and ovarian tissue (Suppl. Fig. 2), histopathologic evaluation of tissue from Animal 1 showed replacement of normal oviduct by numerous endometrial glands (Fig. 1a) within loosely arranged, myxoid or collagenous stroma that expanded the serosa, peri-ovarian tissue and obliterated the oviduct (Fig. 1a, inset). Ectopic endometrial glands often contained blood and cellular debris and were lined by low cuboidal to pseudostratified columnar epithelium that was diffusely and strongly positive for cytokeratin and variably positive for vimentin by IHC (Suppl. Fig. 2). These features are consistent with mixed differentiation endometriosis in rhesus macaques11. Multiple foci within stromal regions of poorly differentiated endometriosis had extensive necrosis (Fig. 1b), with mixed inflammatory infiltrates of macrophages, lymphocytes, and plasma cells, with foci of neutrophils and necrotic cellular debris, and prominent intranuclear inclusions consistent with MPXV (Fig. 1b, inset). Both glands and stroma in regions of endometriosis were diffusely immunoreactive for CD10 (Fig. 1c), a diagnostic marker for human endometriosis12. Mpox immunoreactivity was observed in both glands (Suppl. Fig. 3) and stroma in regions of poorly differentiated endometriosis by IHC (Fig. 1d, inset; e) and correspond to regions with extensive necrosis and mixed inflammatory infiltrates (Fig. 1a, b). Stroma immunoreactive for mpox was characterized by the presence of numerous smooth muscle actin (SMA)-positive vessels (Fig. 1f), extensive vimentin-positive areas (Fig. 1g), many CD68-positive macrophages (Fig. 1h), and few cytokeratin-positive cells (Fig. 1i).

Fig. 1
figure 1

Mpox replication in periovarian glandular and stromal endometriosis in a rhesus macaque. (a) H&E staining of ovary and oviduct from mpox-infected rhesus macaque with endometriosis (Animal 1), (b) higher magnification of inset (a) showing necrosis admixed with macrophages and rare neutrophils with poxviral inclusion (inset, arrow). Serial sections from (a) showing IHC for (c) CD10 and (d) mpox showing mpox in endometrial glands and stroma (box). High magnification of IHC signal in stroma (d, box) for mpox (e) smooth muscle actin (SMA) (f), vimentin (g), CD68 (h), and pan-cytokeratin (CK A/E, i) showing strong positivity for vimentin and CD68 in regions of stromal mpox immunoreactivity. IHC signal = brown or red. Scale bars = 100uM (b, e, f–i); 1 mm (a); 2 mm (c, d).

Full size image

Mpox replication in the endocervix is associated with intraepithelial macrophages

Additional reproductive tract tissues including cervix, uterus, and ovary from both mpox challenged female macaques and prostate and testis from the male mpox challenged macaque were screened by IHC for MPXV. Animal 1 also had evidence of mpox virus in the endocervix (Fig. 2a). No evidence of virus was detected via IHC or in situ hybridization (ISH) in the evaluated tissues of the reproductive organs of Animal 2 (non-endometriosis female) or in the testis and prostate from the male macaque (Table 1). Mpox infection of the endocervix was confirmed by ISH for viral RNA (vRNA) (Fig. 2b, c). Epithelial cells of the cervical glands were focally strongly positive for mpox vRNA, while cervical stroma had rare positivity. Mpox vRNA positive signal corresponded to vimentin immunoreactivity in cervical stroma (Fig. 2d). Mpox vRNA signal in the endocervical glands corresponded to positivity for CD68 (macrophages) by IHC (Fig. 2e). Dual ISH for both mpox and CD68 (macrophages) RNA showed extensive positivity of CD68 in ovarian stroma and few CD68 positive cells in the cervix in the non-endometriotic macaque (Animal 2, Suppl. Fig. 4). While co-localization of mpox in CD68 positive macrophages was observed in Animal 1, the signal for mpox RNA was significantly higher than the CD68 signal (in most cases obscuring visualization) suggesting that while mpox infection was associated with cervicitis in Animal 1, the infected cells were not limited to macrophages but include epithelial and stromal cells as well. To further characterize mpox infected cells in endometriosis, we performed cyclic tissue immunofluorescence (CyCIF) using a targeted panel of pan-cytokeratin, vimentin, and SMA, followed by ISH for MPXV RNA and cytokeratin 10 (CK10) (Fig. 2f–h). Infected cervical epithelial cells were pan-cytokeratin positive while infected cervical stroma was positive for vimentin (Fig. 2g, h). Squamous epithelium of the vagina (CK10) was negative for mpox via ISH (Fig. 2b, f).

Fig. 2
figure 2

IHC and ISH shows mpox replication in endocervical gland macrophages and stroma of the rhesus macaque endocervix. Serial sections of cervix from mpox-infected rhesus macaque with endometriosis (Animal 1) showing (a) endocervical glands and stroma (H&E) positive for (b, c) mpox vRNA by ISH (red), CK10 (teal), (d) vimentin positive stroma, (e) and CD68 positive cervical glands by IHC (brown). (f–h) Cyclic IHC and ISH on cervix (IF: DNA = blue, vimentin = green; SMA = yellow; CK A/E, white; ISH: DNA = blue CK10 = teal, mpox = red). Scale bars = 5 mm (a, b); 100 mM (c–e); f (5 mm); g (200 mM).

Full size image

Discussion

MPXV can productively infect and replicate within diverse cell types, including oral and respiratory epithelium and antigen-presenting immune cells, including monocytes, macrophages, B cells, and dendritic cells2. Early studies of the pathogenesis of high dose aerosolized mpox Zaire (Mpox-Z) in cynomolgus macaques highlighted a role for the mononuclear phagocyte system in virus distribution and reported that mpox infection was associated with ovarian, uterine, and testicular inflammation13. Endometriosis lesions include multiple components of normal uterine tissue, including glandular epithelium and stroma. In Animal 1, regions of glandular epithelium within endometriotic tissue stained strongly for both cytokeratin and vimentin consistent with a de-differentiated state. These same regions were intensely positive for mpox viral protein. Given the strong predilection of mpox for squamous epithelium, it is perhaps not surprising that the endometriotic glandular epithelium would be permissive to mpox replication. More interesting was the prominent mpox replication in the stromal portions of the poorly differentiated endometriosis that were extensively positive for vimentin. It is well-known that Vaccinia virus, a prototypical poxvirus, associates with vimentin intermediate filaments during assembly14. Indeed, we previously reported mpox replication in muscle and adipocytes underlying skin lesions during primary infection with mpox highlighting broad cell tropism for mpox including mesenchymal tissues10. In our study, we did not observe uniform distribution of virus in the reproductive organs across all challenged macaques. Differences in virus15, species of macaque16, and route of infection17 have each individually been documented to result in differences in clinical disease manifestation. All three features differed in the current study, which likely contributed to differences in lesion distribution as compared to previously published reports.

Notably, endometriotic stroma and endocervical glands that were immunopositive for mpox also contained numerous CD68 positive macrophages. Tissue resident antigen-presenting cells such as dendritic cells and macrophages are well-described in the genitourinary tract of people, but their role in the pathogenesis of many sexually transmitted diseases, including mpox, is understudied18 Human mpox patients were often reported to have previous or concurrent infection with other sexually transmitted diseases19. Monocytic inflammation in the genitourinary tract may play a role in susceptibility to mpox. It has also been shown that both proliferative resident macrophages and recruited monocytes are central to the initiation and progression of disease in endometriosis20. MPXV replication has been shown in both human macrophage cell lines21 and in cynomolgus macaque liver Kupffer cells13. Mpox replication in Hofbauer cells, fetal macrophages of the placenta, was reported in a case from a previous outbreak in the Democratic Republic of the Congo22. Recruitment of monocytes to tissues with a predisposition for MPXV cell entry likely promotes enhanced viral replication during mpox infection. Due to the cross-sectional nature of this study, it is unclear to what extent the macrophage infiltration associated with endometriosis contributed to the extensive mpox replication in the endometriotic tissues of Animal 1. Our previous work showed that viral loads in blood peak at 10 days following infection and typically clear by 28 days irrespective of challenge route10, however detailed experimental studies on poxvirus replication and clearance in tissues with the 2022 strain have not been performed. Recently, macrophages were also proposed as one potential source of mpox infection in the brain23. Our finding of macrophage associated mpox replication in reproductive tissues from macaques infected with mpox from the 2022 outbreak, supports the potential role of macrophages in the pathogenesis of other mpox associated clinical syndromes, including neurological disease10. Given that endometriosis has been linked to local immune suppression in the reproductive tract24, it is likely that the combination of macrophage infiltrates and poor viral control due to local immune suppression in this monkey contributed to the increased mpox replication in the reproductive tract despite relatively fewer skin lesions compared to the non-endometriotic animal (Animal 2). Further study on tissue distribution and persistence of MPVX in tissues as compared to plasma would inform whether macrophages or other cell types contribute to prolonged tissue replication in certain clinical settings.

To date, reports of mpox lesions in the reproductive tract of women have been limited to descriptions of vulvar epithelial lesions and a single report of mpox lesions on the external cervical os in women, composed of squamous epithelium25, however, assessing uterine tissue in human patients requires invasive methods that are rarely performed. Two prior studies of cynomolgus macaques (Macaca fascicularis) experimentally infected by aerosolized MPXV reported superficial lesions in the vagina and uterus, with rare necrotic foci in uterine stroma and myometrium13,26. To our knowledge, there are no reports of mpox in endometrial tissue in women or in women affected by endometriosis. This report of MPXV in the endocervix as well as poorly differentiated endometriotic tissue extending to the ovary and fallopian tubes in a mpox infected rhesus macaques suggests that women with endometriosis may be susceptible to similar lesions. Given up to 40% of mpox infected individuals are also HIV positive, associated immune dysfunction likely heightens the risk of developing serious sequelae with viral spread27. In addition, endometriosis has been shown to cause an increase in dysfunctional macrophages within the uterus and peritoneal cavities20, possibly allowing these cellular targets to be more susceptible to mpox infection. Interestingly, the animal with endometriosis had fewer pox lesions overall, suggesting local immune suppression within the endometriotic lesions may have contributed to viral replication. Further study of the prevalence of mpox in women with endometriosis and the pathogenesis of mpox infection in non-epithelial tissues is warranted. It is important for clinicians to be aware of the potential for mpox to spread intraabdominally in high-risk or mpox-exposed women with endometriosis.

Methods

Animal infections

Animal studies were described in10. Briefly, 3 rhesus macaques (Macaca mulatta)—two female, one male – were inoculated intravenously with mpox (MPXV/USA/MA001/2022; lineage B.1, clade 2b; BEI NR-58622; 106 TCID50 (108 plaque-forming unit [PFU]); intravenous). On day 10 post-infection, a full necropsy with description of gross lesions and collection of major organs was performed. All animal studies were approved by the institutional animal care and use committee of Bioqual, Inc. in accordance with the Public Health Service Policy on Humane Care and Use of Laboratory Animals; The Guide for the Care and Use of Laboratory Animals; the U.S. Government Principles for the Utilization and Care of Vertebrate Animals Used in Testing, Research, and Training; and ARRIVE guidelines.

Histopathology

Tissues were fixed in 4% paraformaldehyde for 24 h, transferred to 70% ethanol, and paraffin embedded and blocks sectioned at 5 mm for routine hematoxylin and eosin staining (H&E) and for IHC and ISH. Tissue pathology was independently assessed by three veterinary pathologists (AJM, GMR, CEL).

Immunohistochemistry

Immunohistochemical staining was performed on selected formalin-fixed, paraffin-embedded sections of reproductive tissues using standard techniques. The slides were baked for 30 min at 60ºC, deparaffinized through xylene, 100% ethanol, 95% ethanol, and 1 × tris-buffered saline (TBS). Heat-induced epitope retrieval was performed with a 10% citrate buffer (Sigma-Aldrich, C9999) for all antibodies. To detect mpox, primary mouse anti-Vaccinia antibody (Santa Cruz, SC-58210) was applied at 1:100 followed by mouse Mach-2 HRP-Polymer (Biocare MHRP520) for 30 min, then Nova-Red (Vector, SK4800) for 10 min, and counterstained with hematoxylin followed by bluing using 0.25% ammonia water. For CD10 (1:4000; SinoBiological, 90177-C07H), primary antibody was applied for 60 min by rabbit Mach-2 HRP-Polymer (Biocare, RHRP520) for 30 min, then 3,3’-diaminobenzidine (DAB, Cell Marque 957D-30). The slides were counterstained using 50% hematoxylin solution (BioCare CATHE-MM). Vaccinia and CD10 IHC was performed using a Biocare intelliPATH autostainer. For cytokeratin (1:140, clone AE1/AE3, Dako M3515), CD68 (1:410, clone KP1, Dako M0814), alpha smooth muscle actin (1:1000; clone 1A4, Dako M0851), and vimentin (1:162; clone 3B4, ProGen, 61013) primary antibodies were diluted in Da Vinci Green Diluent (BioCare PD900M) and incubated for 30 min at room temperature, slides were then washed and treated with biotinylated horse anti-mouse secondary antibody (Vector Labs BA-2001, diluted 1:200 in Da Vinci Green) for 30 min at room temperature. Elite Avidin–Biotin Complex solution (Vector Labs, PK-6100) was applied and incubated for 30 min, followed by DAB solution. Matched negative control slides were incubated with Universal Negative Control Serum (BioCare, NC498L). All slides were counterstained using 50% hematoxylin solution (BioCare, CATHE-MM) and were scanned at 20 × using a Midi II Scanner (3DHistotech) on default brightfield settings.

In situ hybridization

Duplex chromogenic in situ hybridization was performed using the RNAscope 2.5 HD Duplex Detection Kit (ACD Bio 322500) with customized and general probes 1260381-C1 (Mmu-KRT10-C1), 504341-C2 (Mmu-CD68-C2), 1226271-C2 (V-MPXV-OPG124-C2), 461341 (Mmu-POLR2a), 457711-C2 (Mmu-PPIB-C2), 2-plex negative control (320751) following recommended guidelines (Protocol 322500-QCK Rev B) as previously described10. Slides were then scanned at 20 × using a Midi II Panoramic Scanner (3D Histotech) on default brightfield settings.

Cyclic ISH and fluorescence microscopy

Cyclic dual RNAscope ISH and immunofluorescence staining was also performed on selected cervical tissues from Animal 1 (T451F). Baking, deparaffinization, and rehydration were performed in the same manner as with IHC, with the substitution of 1 × phosphate-buffered saline with 0.2% fish skin gelatin (FSG, Aurion 900.033) in place of 1 × TBS. Heat-induced epitope retrieval was achieved using the same method as above. The slides were then washed with 1 × PBS/FSG twice for 5 min. A protein block was performed using Intercept Blocking Buffer (LiCor 927-70001) for 30 min. A photochemical bleaching solution (3% H2O2 in 1xPBS with 20 mM NaOH)28 was used prior to antibody application to decrease autofluorescence and between cycles of fluorescence staining. For each application slides were photo bleached for 60 min at room temp while illuminated from above and below with LED light panels (Miroco MI-CL008) on full power. Slides were washed twice with 1 × PBS/FSG for 5 min. In a lightproof humidity chamber, the slides were protein blocked with Intercept for 30 min. For cycle 1, primary antibodies cytokeratin AE1/AE3 (1:140) and SMA (1:100) with Hoechst 1:10,000). For cycle 2, primary antibody vimentin 1:50 with Hoechst 1:10,000 incubated overnight at 4ºC. For both cycles, secondary antibodies against mouse IgG1 conjugated with AlexaFluor 647 (1:1000, Jackson Immunoresearch 115-605-205) and mouse IgG2a conjugated with AlexaFlour 488 (1:1000, JIR 115-545-206) were applied and incubated sequentially at room temperature for 30 min each. All primary and secondary antibodies were diluted in Intercept. After each cycle slides were coverslipped with 80% glycerol solution and scanned at 20 × using a Midi II Scanner in fluorescence mode. Between cycles coverslips were removed by soaking in 1xPBS/FSG for approximately 10 min. For cycle 3, dual-plex RNAscope was performed as listed above followed by detection with the Multiplex-Fluorescent reagent kit V2 (323100) as recommended.

Slide analysis

HALO (v3.6, Indica Labs) was used to fuse serially obtained fluorescence scans with the HALO registration module within the HiPlex FL Module (v4.2). The fused image underwent thresholding for each fluorescent color based on isotype control slides and according to chromogenic IHC identification of staining pattern and intensity.