Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Infection by Intranasal or Intratesticular Route Induces Testicular Damage

Abstract

Background

The role of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in the pathogenesis of testicular damage is uncertain.

Methods

We investigated the virological, pathological, and immunological changes in testes of hamsters challenged by wild-type SARS-CoV-2 and its variants with intranasal or direct testicular inoculation using influenza virus A(H1N1)pdm09 as control.

Results

Besides self-limiting respiratory tract infection, intranasal SARS-CoV-2 challenge caused acute decrease in sperm count, serum testosterone and inhibin B at 4–7 days after infection; and chronic reduction in testicular size and weight, and serum sex hormone at 42–120 days after infection. Acute histopathological damage with worsening degree of testicular inflammation, hemorrhage, necrosis, degeneration of seminiferous tubules, and disruption of orderly spermatogenesis were seen with increasing virus inoculum. Degeneration and death of Sertoli and Leydig cells were found. Although viral loads and SARS-CoV-2 nucleocapsid protein expression were markedly lower in testicular than in lung tissues, direct intratesticular injection of SARS-CoV-2 demonstrated nucleocapsid expressing interstitial cells and epididymal epithelial cells, While intranasal or intratesticular challenge by A(H1N1)pdm09 control showed no testicular infection or damage. From 7 to 120 days after infection, degeneration and apoptosis of seminiferous tubules, immune complex deposition, and depletion of spermatogenic cell and spermatozoa persisted. Intranasal challenge with Omicron and Delta variants could also induce similar testicular changes. This testicular damage can be prevented by vaccination.

Conclusions

SARS-CoV-2 can cause acute testicular damage with subsequent chronic asymmetric testicular atrophy and associated hormonal changes despite a self-limiting pneumonia in hamsters. Awareness of possible hypogonadism and subfertility is important in managing convalescent coronavirus disease 2019 in men.

hamster, coronavirus, COVID-19, SARS-CoV-2, testicle

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has caused the devastating coronavirus disease 2019 (COVID-19) pandemic [1]. Although COVID-19 is primarily a respiratory infection, involvements of many extrapulmonary tissues from olfactory sensory system to reproductive tract have been reported [2, 3]. COVID-19–associated testicular pain was reported to be more frequent than expected [4, 5]. In a human challenge study, 1 in 18 SARS-CoV-2–infected volunteers experienced epididymal discomfort [6]. One postmortem study of COVID-19–infected men revealed orchitis with extensive germ cell and Leydig cell destruction [7]. Semen specimens from 4 patients with acute COVID-19 and 2 recovered from COVID-19 were SARS-CoV-2 positive by reverse-transcription polymerase chain reaction though this was not consistently found in recovered patients [8, 9]. Whether SARS-CoV-2 can cause testicular damage in COVID-19 is still uncertain. We investigated the temporal profiles of virological, pathological, immunological, and hormonal changes in hamsters challenged by intranasal or direct testicular administration of wild-type and variant SARS-CoV-2, using A(H1N1)pdm09 influenza virus as a control.

METHODS

Animal, Virus, and Biosafety

Male Syrian hamsters, 8–10 weeks old, were infected with SARS-CoV-2 wild-type strain HK-13 (GenBank accession no. MT835140), and Delta (GISAID: EPI-ISL-3221329) and Omicron (GISAID: EPI-ISL-7138045) variants [10]. A(H1N1)pdm09 (A/HK/415742/2009) influenza mouse-adapted strain was used as control [11]. Experiments involving live SARS-CoV-2 followed the approved standard operating procedures of the biosafety level 3 facility [12, 13]. Experimental details are included in the Supplementary Methods. The animal experiments were approved by the HKU Committee on the Use of Live Animals in Teaching and Research and complied with the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines. Male Syrian hamsters were obtained from the Chinese University of Hong Kong Laboratory Animal Service Centre through the HKU Centre for Comparative Medicine Research

Intranasal Virus Inoculation and Vaccination

Hamsters were anesthetized and challenged intranasally with 10–10⁵ plaque-forming unit (PFU) of SARS-CoV-2 strains or 10³ or 10⁵ PFU of A(H1N1)pdm09 in 50 µL volumes [12]. The animals were euthanized for analysis at 1, 4, 7, 42, or 120 days post infection (dpi), together with mock-infected controls. Formalin-inactivated SARS-CoV-2 HK-13 whole-virion vaccine (10 µg of total protein per hamster) was given in a testicular protection experiment using phosphate-buffered saline (PBS) as the control [14].

Intratesticular Virus Inoculation

In anesthetized hamsters, SARS-CoV-2 HK-13 or A(H1N1)pdm09 (10⁵ PFU in 50 µL) was injected transcutaneously into the testes with a 27-gauge needle. The mock-infected controls were injected with 50 µL of PBS. The animals were euthanized at 1 or 4 dpi.

Testis Weight and Sperm Count

At the time of sampling, testes were dissected from the epididymis and weighed. Freshly excised epididymis was incised in 1 mL of PBS. Sperm were collected after 15 minutes and counted in a hemocytometer.

Viral Load Assays, Histopathological Examination, and Cytokine and Chemokine Profiling

Testicular viral loads, histopathology, immunohistochemistry, apoptosis marker, and cytokine/chemokine profiles, were performed as described previously [11, 12, 15]. Detailed protocols are provided in the Supplementary Methods. Specific primers and probes for SARS-CoV-2 RNA-dependent RNA polymerase gene, influenza matrix gene, and cytokines/chemokines for real-time polymerase chain reaction assay are listed in Supplementary Table 1.

Paraffin sections of testes were stained with hematoxylin-eosin. Biomarkers for seminiferous tubular germ cells and Sertoli cells were stained by immunohistochemical method with antibodies to “deleted in azoospermia-like” (DAZL) protein and vimentin, respectively.

SARS-CoV-2 nucleocapsid protein (N protein) or influenza viral nucleoprotein (NP) expression in testes were stained with a rabbit antibody against SARS-CoV-2 N protein [12] or mouse anti–influenza nucleoprotein antibody [11, 16]. Histological examination was conducted in a blinded manner. The damages in testicular tissue were scored using Johnsen's score criteria with minor modification (Supplementary Table 2) [17, 18]. Apoptosis was detected using the Click-iT Plus TUNEL (terminal deoxynucleotidyl transferase dUTP nick-end labeling) assay kit and was scored semiquantitatively [10].

Enzyme Immunoassay or Male Sex Hormone

Morning serum total testosterone and inhibin B were detected with hamster enzyme immunoassay Kits (MyBioSource).

Statistical Analysis

Statistical analyses for significant differences (P < .05) were performed with Student's t-test or Fisher's exact test to compare infection and mock-infected control groups (GraphPad Prism8 software).

RESULTS

Acute Testicular Damage Caused by Intranasal SARS-CoV-2 Challenge in Hamsters

Male hamsters were intranasally inoculated with 10, 10², 10³, or 10⁵ PFU of wild-type SARS-CoV-2 (Figure 1A). Self-limiting histological changes of pneumonia were similar to what our group previously reported [12]. No scrotal swelling was clinically detectable in any animal at any time point. Sperm counts were significantly reduced in all hamsters at 4 days after 10³ or 10⁵ PFU infection (Figure 1B), despite absence of gross pathological changes in appearance, size, or weight (Figure 1C and 1D). Serum testosterone and inhibin B levels were significantly lower at 4 and 7 dpi compared with the mock-infected control (Figure 1E). The findings are consistent with subclinical testicular damage during the acute phase of SARS-CoV-2 infection.

Figure 1.

Intranasal inoculation of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causes acute testicular damage in golden Syrian hamsters. A, Schema of intranasal challenge of male Syrian hamsters at 8–10 weeks of age. The animals were randomly divided into groups and intranasally challenged with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) HK-13 strain at a dose of 10, 10², 10³, or 10⁵ plaque-forming units (PFU) in 50 µL. They were monitored and euthanized at 1, 4, or 7 days post infection (dpi) for study of acute damage or at 42 or 120 dpi for chronic damage. B, Sperm counts. The epididymis was dissected and opened, and sperm were rinsed out to count the number per hamster testicle (n = 5–10 hamsters per group). C, Average weight of testis. After removing the epididymis, the wet weight of the testis was measured immediately (n = 5–10 per group). D, Representative images of gross pathology in testicles from the SARS-CoV-2–infected group and mock-infected controls. The images show no apparent changes in size and color at 1, 4, or 7 dpi compared with controls. Scale bar represents 1 cm. E, Serum concentration of sex hormones determined with enzyme-linked immunosorbent assay. Serum samples were collected in the mornings for assays of testosterone and inhibin B 1, 4, or 7 days after inoculation of 10³ or 10⁵ PFU of SARS-CoV-2 HK-13 strain (n = 5–13 per group). B, C, E, Data represent means with standard deviations. *P < .05; **P < .01 (Student’s t-test for comparison with controls).

Chronic Testicular Damage After Intranasal SARS-CoV-2 Challenge

Grossly reduced size and weight of testes were observed at 120 dpi (Figure 2A and 2B). Sperm counts were significantly reduced at 42 and 120 dpi (Figure 2C), as were serum testosterone and inhibin B levels at 120 dpi (Figure 2D). Table 1 summarizes findings at different time points after 10³ PFU infection, which indicates that testicular and hormonal changes continued to evolve despite the lungs have recovered from SARS-CoV-2 infection.

Table 1.

Testicular Damage Caused by Severe Acute Respiratory Syndrome Coronavirus 2 in Syrian Hamsters^a

Variable	Mock-Infected Controls	Acute Damage		Chronic Damage
		4 dpi	7 dpi	42 dpi	120 dpi
No. of hamsters	14	17	6	7	7
Testes weight, g	2.06 (0.14)	1.81 (0.38)	1.9 (0.27)	1.87 (0.42)	1.39 (0.30)
Sperm count, ×107/side	14.71 (5.00)	8.02 (4.17)	7.81 (5.78)	5.85 (3.97)	2.50 (3.27)
Hamsters with testicular damage, no./total (%)	1/14 (7.14)	8/17 (47.06)	4/6 (66.67)	4/7 (57.14)	5/7 (71.43)
Johnsen score for seminiferous tubule damage	8.87 (1.53) (n = 6)	6.16 (1.39) (n = 8)	4.51 (2.25)	6.44 (0.51)	5.39 (1.60)
TUNEL score for apoptosis	1 (0) (n = 8)	2.6 (0.55) (n = 5)	2.67 (1.15) (n = 3)	2.67 (0.58) (n = 3)	2.8 (0.84) (n = 5)
Serum testosterone, ng/L	822.97 (87.39)	591.69 (251.91)	473.04 (388.82)	762.3 (23.19)	399.17 (387.67)
Serum inhibin-B, pg/mL	194.83 (61.99)	114.55 (77.35)	90.33 (59.58)	69.75 (79.33)	66.69 (53.31)

Abbreviations: dpi, days post infection; TUNEL, terminal deoxynucleotidyl transferase dUTP nick-end labeling.

Severe acute respiratory syndrome coronavirus 2 was given at a dose of 10³ plaque-forming units. Values within table represent mean (standard deviation) unless otherwise specified.

*P < .05; **P < .01; ****P < .0001 (comparison with mock-infected controls; Student t test).

P < .05; ^††P< .01

Figure 2.

Intranasal severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) challenge causes chronic testicular damage in golden Syrian hamsters. Male hamsters were intranasally inoculated with 10³ plaque-forming units (PFU) of wild-type SARS-CoV-2 HK-13 stain in 50 µL. The animals were monitored and scarified at 42 or 120 days post infection (dpi). A, Representative images of gross pathology of testicles after SARS-CoV-2 infection and age-matched mock-infected controls. The images show reduced size of testes at 120 dpi, with no apparent changes at 42 dpi compared with controls. Scale bar represents 1 cm. B, Average weight. After removal of the epididymis, the testis was immediately measured for wet weight (n = 5–7 per group). C, Sperm counts. The epididymis was dissected and the number of sperm was counted per hamster testicle (n = 5–7 per group). D, Serum concentration of testosterone and inhibin B determined with enzyme-linked immunosorbent assay (n = 4–7 per group). B–D, Data represent means with standard deviations. **P < .01; ***P < .001 (Student’s t-test for comparison with controls).

Histopathological Changes of Acute Testicular Damage in SARS-CoV-2 Infection

The normal structure of testes in mock-infected control hamsters showed seminiferous tubules consisting of multilayers of germline epithelial cells neatly arranged with mature spermatozoa in the lumen (Figure 3A). At 1 dpi, intranasal challenge with 105 PFU, 33% of hamsters (2 of 6) had expanded testicular interstitial space owing to edema. Although the multilayer structure remained intact, mild segmental seminiferous epithelial degeneration and germ cell sloughing were observed (Supplementary Figure 1). At 4 dpi, severe testicular hemorrhage and interstitial mononuclear cell infiltration were observed in 18.7% of hamsters (3 of 16) (Figure 3B). Severe seminiferous tubular necrosis with occasional neutrophils and disordered germ cells arrangement with reduced layers of spermatogenic cell spectrum were found in 43.8% (7 of 16). Multinucleated germ cell degeneration and even depletion of luminal spermatozoa were frequently observed (Figure 3C). Of the hamsters infected with 10³ PFU, 47.06% (8 of 17) showed testicular damage at 4 dpi, which was generally less severe and more patchy than in the 10⁵ PFU group.

Figure 3.

Histopathological changes in the testicles of golden Syrian hamsters challenged intranasally with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) at 4 or 7 days post infection (dpi). SARS-CoV-2 HK-13 strain was intranasally inoculated at a dose of 10, 10², 10³, or 10⁵ plaque-forming units (PFU) in 50 µL per animal. Testicles were harvested at 4 or 7 dpi and immediately fixed in neutral formalin. Paraffin sections were stained by hematoxylin-eosin (HE) for microscopic examination. A, Illustration (left) and representative images of HE-stained testicular section of mock-infected control showing neatly arranged testicular structure, seminiferous tubules and different cell types, including Sertoli cells and germ cells at various differentiation stages with spermatozoa in the lumen of seminiferous tubules. Scale bars represent 500, 100, or 50 µm. B, Representative HE-stained images of testicular tissues from hamsters intranasally challenged with 10⁵ PFU of SARS-CoV-2 at 4 dpi. Interstitial inflammatory damage included hemorrhage, exudation with cellular infiltrates, and seminiferous tubule luminal edema (asterisk) are shown in the left 2 images. Extensive interstitial and seminiferous tubular immune cell infiltration are shown in the images on the right; the boxed area was magnified for showing mononuclear cell infiltrates. Scale bars represent 100 µm. C, Representative HE-stained images of testicular tissues from hamsters intranasally challenged with 10⁵ PFU of SARS-CoV-2 at 4 dpi show diffusely distributed abnormal seminiferous tubules and enlarged lumen without spermatozoa (asterisk), reduced germinal cell spectrum (layers), and disordered cellular arrangement in most seminiferous segments (solid arrows). Degeneration of germ cells forming multinucleated giant cells is indicated by thin arrows. Severe seminiferous tubular necrosis (double asterisks), with luminal cells debris and infiltrated polymorph-nucleated cells in the magnified boxed area, is indicative of neutrophil infiltration. Scale bars represent 100 or 50 µm. D, Representative HE-stained images of testicular tissues from hamsters intranasally challenged with 10⁵ PFU of SARS-CoV-2, at 7 dpi. The left 2 images show inflammatory infiltration and edema in the interstitium and lumen of expanded seminiferous tubules (asterisk). Severe diffuse seminiferous tubular epithelial necrosis with cell debris filled the lumen. Spermatogenic cells were loosened and sloughed (2 images on right). Scale bars represent 200 or 100 µm. E, Scores for the spermatogenic epithelium. The integrity of germinal epithelium in seminiferous tubules was semiquantitatively assessed based on Johnsen scores, with minor modifications. Data represent means with standard deviations (SDs). **P < .01 (Student's t-test for comparison with mock-infected controls). F, Representative HE-stained images of the epididymis of mock-infected controls or hamsters challenged with 10⁵ PFU of SARS-CoV-2 at 4 and 7 dpi. The epididymal lumens of controls were filled with mature sperm, while those of infected hamsters contained large amounts of sloughed germ cells and cell debris with much-reduced sperm counts. Interstitial mononuclear immune cells infiltration was found in tissues at both 4 and 7 dpi. Scale bars represent 200 or 100 µm.

At 7 dpi, severe testicular necrosis was observed in both 10³ PFU– and 10⁵ PFU–infected hamsters. Some seminiferous tubular lumens contained only cell debris (Figure 3D). Significantly lower Johnsen's scores at 4 and 7 dpi (Figure 3E and Table 1) indicated substantial impairment of spermatogenesis. The epididymis had interstitial mononuclear cell infiltration with lumens filled with sloughed germ cells and cell debris (Figure 3F). No detectable histopathological changes were observed at 4 dpi, after challenge with 10 or 10² PFU (Supplementary Figure 1).

Degeneration and Necrosis of Sertoli Cells and Spermatogenic Cells After Intranasal SARS-CoV-2 Infection

Sertoli cells with long cytoplasmic arms extending toward the lumen situated at the basal membrane of the seminiferous tubules are supportive cells that nourish and support orderly germ cell differentiation. Sertoli cells expressing high levels of cytoplasmic vimentin were regularly distributed as a radiant pattern from the basal membrane toward the lumen (Figure 4A). Severely disrupted distribution pattern of vimentin-expressing Sertoli cells was seen at 4 days after infection with 10⁵ PFU (Figure 4B). Cytoplasmic vacuolation, degeneration and detachment of Sertoli cells into the lumen were also found (Figure 4B).

Figure 4.

Immunohistochemical staining of Sertoli cells and spermatogenic cells in testicular tissue of golden Syrian hamster challenged by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) at 4 days post infection (dpi). A, Representative images of Sertoli cell biomarker vimentin in testis of mock-infected control. The images show normal seminiferous tubules (left) and the Sertoli cells with cytoplasmic vimentin regularly distributed in a radiant pattern from the basal membrane, their cytoplasmic arms extending toward the lumen (open arrows). Scale bars represent 10 or 50 µm. B, Representative images of testes tissue of hamsters intranasally challenged by 10⁵ plaque-forming unit (PFU) of SARS-CoV-2 at 4 dpi. Vimentin-expressing Sertoli cells were deformed and detached from basal membrane (left and magnified square area). Images on the right show Sertoli cells cytoplasmic vacuolation degeneration and loss of cytoplasmic arms (open arrows in magnified square area). Scale bar represents 50 µm. C, Representative images of deleted in azoospermia-like (DAZL) protein in testicular section of controls. The low-magnification images show that DAZL was highly expressed and evenly distributed in seminiferous tubular epithelium; higher-power image in the middle and magnified square area show extensive and high-level expression of DAZL in primary spermatocytes (solid arrows) and reduced level in spermatogonia (arrowheads), while the secondary spermatocytes were negative (thin arrows). Open arrows indicate Sertoli cells, which were negative for DAZL. Scale bars represent 100 or 50 µm. D, Representative images of DAZL expression at 4 dpi in testes of hamsters intranasally challenged by 10⁵ PFU of SARS-CoV-2. The left 2 images show that DAZL-expressing germinal cells were disarranged, detached, or in the form of multinucleated giant spermatocytes (solid arrows), spermatogonia was absent from the 2 severely damaged seminiferous tubules (areas within dashed lines), and the thin arrows indicate multinucleated giant cells, which are DAZL negative, suggestive of degenerated secondary spermatocytes. The image on the right and magnified square area shows that the sloughed germ cells were both DAZL positive and DAZL negative. Scale bars represent 100 or 50 µm. E, Images of hematoxylin-eosin (HE)–stained testicular tissues showing the morphology of interstitial Leydig cells. The Leydig cells were found in the interstitial space of control testicular section (left image and magnified square area; open arrowheads). At 4 dpi, testis of hamster intranasally challenged by SARS-CoV-2 shows 2 Leydig cells with condensed nuclei indicative of apoptosis (right image and magnified square area; open arrowheads). Scale bars represent 100 or 50 µm.

The DAZL protein was generally more readily expressed in primary spermatocytes than spermatogonia but not in secondary spermatocytes and Sertoli cells (Figure 4C). In hamsters intranasally challenged 10⁵ PFU, at 4 dpi, DAZL staining demonstrated disarray of germ cell arrangement, and DAZL-positive multinucleated giant cells formed from primary spermatocytes, together with many DAZL-negative sloughed germ cells and multinucleated giant cells formed from secondary spermatocytes (Figure 4D). Affected primary spermatocytes and secondary spermatocytes were accompanied by ballooning changes in adjacent Sertoli cells. Interstitial Leydig cells showed nuclear condensation indicative of apoptosis (Figure 4E).

Testicular Infection by Intranasal or Direct Intratesticular Injection With SARS-CoV-2

Unlike highly infected lung tissues, low viral loads were found in only a few testicular samples at 4, 7, 42, and 120 dpi (Figure 5A). Immunofluorescence staining of viral N protein showed only a few positive cells in seminiferous tubules and detached cells in epididymal lumen at 4 dpi (Figure 5B). Sperm smear also showed a few N-positive sloughed cells (Figure 5C).

Figure 5.

Viral load and viral antigen expression in testicles after intranasal or intratesticular inoculation of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). A, Viral load in homogenized testicular tissue and lungs were determined using quantitative reverse-transcription polymerase chain reaction (qRT-PCR). Data are presented as copies of SARS-CoV-2 RNA-dependent RNA polymerase (RdRp) gene per copy of β-actin in log scale. Data represent means with standard deviations (SDs) (n = 5–15 per group). Horizontal dashed line indicates the cutoff for testicular viral load. B, Representative images of immunofluorescence-stained viral nucleocapsid protein (N protein) in the testes (upper panel) and epididymis (lower panel) of hamsters at 4 days post infection (dpi), after intranasal challenge with 10³ plaque-forming units (PFU) of SARS-CoV-2. SARS-CoV-2 N protein–positive cells (green) are indicated by arrows in original and magnified images. Scale bars represent 100 or 50 µm. C, Representative images of immunofluorescent viral N protein in sperm sample smear slides prepared at 4 dpi from hamsters intranasally challenged with 10³ PFU of SARS-CoV-2. SARS-CoV-2 N protein–positive (green) positive cells (sloughed cells) are indicated by arrows in original and magnified images; scale bars represent 50 µm. D, Representative hematoxylin-eosin (HE)–stained images and immunofluorescent viral N protein in the testes of hamsters at 1 or 4 days after intratesticular injection of 10⁵ PFU of SARS-CoV-2 in 50 µL. The HE-stained images show interstitial congestion at 1 dpi and mild seminiferous tubular degeneration and interstitial immune cells infiltration at 4 dpi (black thin arrows). Immunofluorescence-stained SARS-CoV-2 viral N protein–expressing cells were found in the interstitium of the testes at 1 and 4 dpi (arrows), and SARS-CoV-2 viral N protein was also found in the seminiferous epithelium, which was suggestive of germ cells. Representative N protein–expressing cells are magnified in the inserts. Scale bars represent 200, 100, or 50 µm. E, Representative images of immunofluorescence-stained viral N protein in the epididymis at 1 day after intratesticular injection. Images on the left show N protein–expressing epithelial cells of epididymis (arrows in magnified image). Images on the right show N protein–positive spermatocytes in the lumen of epididymis. Scale bar represents 50 µm. F, Viral load in the testes determined by qRT-PCR at 1 or 4 days after intratesticular injection of SARS-CoV-2. Data represent means with SDs (n = 4–6 per group). G, Determination of messenger RNA (mRNA) expression of inflammatory cytokine/chemokine by qRT-PCR in the homogenized testes at 1 or 4 days after injection of SARS-CoV-2 or the same dose of influenza A(H1N1)pdm09 virus (10⁵ PFU), which was injected in another group of hamsters as controls (n = 3–6 per group); data represent means with SDs. *P < .05; **P < .01; ****P < .0001 (Student's t-test; comparison with mock-infected controls). Abbreviations: DAPI, 4ʹ,6-diamidino-2-phenylindole; IFN, interferon; IL-1β, interleukin 1β; IL-6, interleukin 6; sgRNA, subgenomic RNA; TNF, tumor necrosis factor.

Direct intratesticular injection induced localized interstitial inflammatory infiltration and segmental seminiferous tubular degeneration at 4 dpi (Figure 5D). But N protein was readily seen in the interstitium and epididymis epithelial cells as multiple foci with each containing 3–5 immunofluorescence-positive cells at 1 and 4 dpi (Figure 5D and 5E). As 5 of 6 testes had detectable viral loads while 2 had subgenomic RNA expression, SARS-CoV-2 has likely replicated in testicular cells (Figure 5F). Significantly increased cytokine/chemokine messenger RNA (mRNA) in testes was detected after intratesticular injection of SARS-CoV-2 at 1 and 4 dpi (Figure 5G). In contrast to SARS-CoV-2–injected hamsters, influenza viral antigen was rarely found in testes after A(H1N1)pdm09 injection, and no histopathological changes were observed despite marginally detectable viral load and increased mRNA expression for cytokines/chemokines (Supplementary Figure 2). Together, our results reveal that testicular cells can be infected when exposed to SARS-CoV-2 which triggers cytokine/chemokine responses and cell damage.

Inflammatory Cytokine/Chemokine mRNA Expression Profiles and Apoptosis Regulatory Genes in Testes of Hamsters Challenged Intranasally

Because testicular damage developed in only some hamsters (Table 1), only their testicular cytokine/chemokine levels were compared with those mock-infected controls. Levels of proinflammatory cytokines/chemokines, including interferon (IFN) α, IFN-γ, interlekin 6 (IL-6), tumor necrosis factor (TNF) α C-X-C motif chemokine ligand 10 (CXCL10), C-C Motif Chemokine Ligand 3 (CCL3) (macrophage inflammatory protein 1α), and C-C Motif Chemokine Ligand 5 (CCL5) (RANTES [regulated on activation of normal T cells expressed and secreted]) were significantly higher in 10⁵ PFU–infected hamsters (Figure 6A).

Figure 6.

Messenger RNA (mRNA) expression of inflammatory cytokine/chemokine and apoptosis regulatory genes and TUNEL (terminal deoxynucleotidyl transferase dUTP nick-end labeling) staining of testicular tissues from severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infected hamsters or mock-infection control animals. Testes collected at 4 days post infection (dpi), after intranasal challenge by 10³ or 10⁵ plaque-forming unit (PFU) of SARS-CoV-2 were homogenized for total RNA extraction and quantitative reverse-transcription polymerase chain reaction (qRT-PCR). The testes of infected animals showing various degrees of histopathological damage compared with controls were included in this assay to compare cytokine/chemokine gene expression with that in controls. A, Inflammatory cytokine and chemokine mRNA expression level in homogenized testes were determined by qRT-PCR with each gene-specific primer. Housekeeping gene β-actin was included for normalization of RNA concentration in each sample. Levels of relative gene expression were calculated by the Delta-Delta Ct method (ΔΔCt) method relative to controls (n = 7–9 per group). B, Apoptosis regulatory gene expression levels in homogenized testes at 4 dpi, after different doses of SARS-CoV-2 challenge. The relative mRNA expression was determined by qRT-PCR and calculated with the ΔΔCt method (n = 7–10 per group). C, Representative images of apoptosis marker TUNEL-labeled testicular sections from 10³ or 10⁵ PFU of SARS-CoV-2 infected hamsters at 4 dpi or controls. In controls, 1 or 2 TUNEL-positive cells were seen. Small foci of TUNEL-positive cells were seen in 1 seminiferous tubule after infection by 10³ PFU of SARS-CoV-2. Many TUNEL-positive cells were shown in multiple seminiferous tubules after 10⁵ PFU infection. Scale bar represents 200 µm. D, Semiquantitative scores for percentage TUNEL-positive seminiferous tubules. The testicular sections were labeled with the Click-iT Plus TUNEL assay kit and examined under a fluorescence microscope. The TUNEL-positive number of seminiferous tubules were semiquantified from 4 random ×100 magnification fields. TUNEL scores were defined as follows: 0, TUNEL-positive seminiferous tubules in <5% of total seminiferous tubules examined; 1, 5%–25%; 2, 25%–50%; 3, 50%–75%; and 4, >75% (n = 3–7 per group). A, B, D, Data represent means with standard deviations. *P < .05; ** P < .01; ***P < .001 (Student's t-test). Abbreviations: BAD, BCL2-associated agonist of cell death; BCL2, B-cell lymphoma 2; CCL3, C-C motif chemokine ligand 3 (or macrophage inflammatory protein 1α); CCL5, C-C motif chemokine ligand 5 ; CXCL10, C-X-C motif chemokine ligand 10; DAPI, 4ʹ,6-diamidino-2-phenylindole; DR6, death receptor 6; FasR, Fas receptor; IFN, interferon; IL-1β, interleukin 1β; IL-6, interleukin 6; NOXA, phorbol-12-myristate-13-acetate–induced protein 1; PUMA, p53 up-regulated modulator of apoptosis; TNF, tumor necrosis factor; TNFRSF1A, TNF receptor 1A; TRAIL, TNF-related apoptosis-inducing ligand.

The expression of death receptor TRAIL (TNF-related apoptosis-inducing ligand) and proapoptotic gene PUMA (p53 up-regulated modulator of apoptosis) were significantly up-regulated in 10³- and 10⁵ PFU challenged hamsters, respectively (Figure 6B). This suggested that inflammatory cytokine responses were associated with testicular pathology. The abundance of TUNEL staining signal varied; TUNEL-positive cells were minimal in controls, while they were found in multiple seminiferous tubules after SARS-CoV-2 infection. The semiquantitative TUNEL-positive scores increased with increasingly higher virus inoculum at 4 dpi (Figure 6C and 6D).

Histopathological Changes of Chronic Testicular Damage After SARS-CoV-2 Infection

At 42 and 120 dpi, after intranasal 10³ PFU challenge, various degrees of seminiferous epithelial degeneration persisted, including disorganized germ cells, sloughed cells, seminiferous tubular necrosis filled with luminal cell debris, severe spermatogenic cell depletion, and loss of spermatozoa (Figure 7A and 7B). Johnsen scores indicated severe impairment of spermatogenesis (Figure 3E). At 42 and 120 dpi, testicular histopathological damage was seen in 57.1% (4 of 7) and 71.4% (5 of 7) of hamsters, respectively (Table 1). The epididymal lumen of epididymis at 42 or 120 dpi contained few or no sperm, or with just cell debris.

Figure 7.

Chronic testicular damage at 42 or 120 days post infection (dpi) in hamsters infected with 10³ plaque-forming units (PFU) of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). A, Representative images of hematoxylin-eosin (HE)–stained tissue sections of testes and epididymis at 42 dpi. The testes showed severe diffuse seminiferous tubular epithelial necrosis; spermatogenic cells were depleted in enlarged tubular lumen, which contained small numbers of spermatozoa. The epididymal lumen contained a large amount of cell debris. Scale bars represent 200 or 100 µm. B, Representative images of HE-stained testes at 120 dpi. The germinal epithelium of seminiferous tubules show atrophy and loss of normal differentiation spectrum, and the epididymal lumen shows no sperm. Scale bars represent 200 or 100 µm. C, Messenger RNA (mRNA) expression of cytokine/chemokine and apoptosis regulating gene at 42 and 120 dpi compared with expression at the acute stage (4 and 7 dpi), after the same infective doses of SARS-CoV-2 (10³ PFU) (n = 6–8 per group). Data represent means with standard deviations (SDs). Red curves indicate the mean expression level at each time point after infection. D, Representative images of TUNEL (terminal deoxynucleotidyl transferase dUTP nick-end labeling)–stained testicular sections. The mock-infected control testis showed only 1 TUNEL-positive cell, while significantly more abundant TUNEL-positive cells were seen in the sections obtained at 42 and 120 dpi. Scale bar represents 200 µm. E, Scores for the intensity of TUNEL-positive cells. The testicular tissue sections were labeled with the Click-iT Plus TUNEL assay kit and examined under a fluorescence microscope. The TUNEL-positive seminiferous tubules were semiquantified from 4 random ×100 magnification fields. TUNEL scores were defined as follows: 0, TUNEL-positive seminiferous tubules <5% of total seminiferous tubules examined; 1, 5%–25%; 2, 25%–50%; 3, 50%–75%; and 4, >75% (n = 3–7 per group). Data represent means with SDs. ****P < .0001 (Student's t-test). Abbreviations: BAD, BCL2-associated agonist of cell death; BCL2, B-cell lymphoma 2; BAX, BCL-2-associated X protein; BID, BH3 interacting-domain death agonist; DAPI, 4ʹ,6-diamidino-2-phenylindole; DR6, death receptor 6; FasR, Fas receptor; NOXA, phorbol-12-myristate-13-acetate–induced protein 1; PUMA, p53 up-regulated modulator of apoptosis; TNFRSF1A, tumor necrosis factor receptor 1A; TRAIL, tumor necrosis factor–related apoptosis-inducing ligand.

Unexpectedly, mRNA levels of inflammatory cytokine/chemokine at 42 and 120 dpi were similar to those in mock-infected controls (Figure 7C). However, expression of death receptor gene TRAIL, and proapoptosis gene BAX were still elevated at 120 dpi, suggesting progressive pathology. TUNEL staining showed more extensive positive signals in seminiferous tubules at 120 dpi than at 42 dpi (Figure 7D and 7E). Immunofluorescence staining by mouse anti-hamster immunoglobulin G (IgG) showed extensive staining on cell surfaces of necrotic seminiferous tubules at 7–120 dpi, with patches of positive complement component 3 (C3), suggesting complement fixation by immune complex deposits at 7 dpi (Figure 8A–8F). These findings further suggested that the blood-testicular barrier was breached [19, 20], and immune complex deposition might cause the testicular damage.

Figure 8.

Immunofluorescence-stained hamster immunoglobulin G (IgG) and complement component 3 (C3) in testicular tissues. Paraffin-embedded testes samples collected at 1, 4, 7, 42, and 120 days post infection (dpi), after intranasal inoculation with 10³ or 10⁵ plaque-forming units (PFU) of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), were stained for hamster IgG or C3, using goat anti-hamster IgG or C3 antibodies, respectively. A, Representative images of immunofluorescence staining of hamster IgG in testicular tissues from mock-infected control. IgG staining signal was found only in the interstitium (thin arrows). No seminiferous tubular epithelial cells were positively stained. The area encircled by the dashed line was magnified, showing germ cells were negative for IgG. Scale bars represent 200 or 100 µm. B, Representative images of IgG in testes at 1 and 4 dpi, after intranasal challenge with 10⁵ PFU of SARS-CoV-2. The IgG staining signals were found in interstitium. In the testes at 4 dpi, the interstitial IgG was more intensely stained, and seminiferous tubules within the dashed circle show germ cell sloughing but no IgG-labeled germ cells (solid arrows in magnified image). Scale bars represent 200 or 100 µm. C, Representative images of IgG in testes 7 days after SARS-CoV-2 intranasal challenge. IgG staining signal was seen in the interstitium (thin arrows), and extensive IgG signal was found inside seminiferous tubules, where the germ cells were largely depleted (open arrows). In the seminiferous tubules with epithelium destruction (circled areas), IgG labeled the surface of germ cells that were disarranged or sloughed in the lumen of seminiferous tubules (solid arrows). Scale bars represent 200 or 100 µm. D, E, Representative images of IgG in the testes of hamster at 42 (D) or 120 (E) dpi SARS-CoV-2 intranasal challenge. The IgG staining intensity is reduced compared with 7 dpi. IgG was located mostly in the interstitium (thin arrows). IgG was occasionally found precipitated in the seminiferous tubules and on the surface of sloughing germ cells (solid arrows in magnified circled area). Scale bars represent 200 or 100 µm. F, Colocalization of complement C3 with IgG in hamster testicular tissues at 7 dpi, after SARS-CoV-2 intranasal challenge. Complement C3 was occasionally found on the surface of germ cells in the same seminiferous tubule (circled areas) where IgG-labeled germ cells were located. Scale bars represent 200 or 100 µm. Abbreviation: DAPI, 4ʹ,6-diamidino-2-phenylindole.

Testicular Damage After Intranasal Challenge With Multiple SARS-CoV-2 Variant Strains but Not With A(H1N1)pdm09

To determine whether testicular damage is SARS-CoV-2 strain specific, we intranasally challenged groups of 3 hamsters with 10³ PFU of Delta (B.1.617.2) and Omicron (B1.1.529) variants. At 4 dpi, pneumonia occurred in all 6 hamsters with bronchiolitis, alveolitis, and endotheliitis (Supplementary Figure 3), while the severity was less in Omicron-infected hamsters. Degeneration and necrosis of seminiferous tubules with varying severity was observed in the testes of 2 of 3 Delta-infected and 2 of 3 Omicron-infected hamsters both at 4 and 7 dpi (Figure 9A). Possibly owing to more inflammatory edema at 4 dpi, Omicron-infected testes were significantly heavier than those in mock-infected controls but became significantly lighter at 7 dpi (Figure 9B). Sperm counts from Delta- and Omicron-infected hamsters were significantly reduced at 4 and 7 dpi, respectively (Figure 9C).

Figure 9.

Multiple severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variant strains but not (H1N1)pdm09 influenza virus can cause testicular damage in hamsters. For each SARS-CoV-2 variant strain, including Delta (B.1.617.2) and Omicron (B.1.1.529), 10³ plaque-forming unit (PFU) were intranasally inoculated into 6 hamsters per variant of concern (VOC). Influenza A pdm(H1N1)2009 virus infection was performed as a control. The animals were scarified at 4 and 7 dpi. Lung, testicles, and blood were taken for analysis. A, Representative hematoxylin-eosin (HE)–stained images of testes at 4 and 7 days post infection (dpi), after intranasal inoculation with 10³ PFU of Delta (B.1.617.2) or Omicron (B.1.1.529) variants. Various degrees of seminiferous tubular degeneration were shown. Scale bar represents 200 µm. B, Testis weight at 4 and 7 days after Omicron or Delta virus challenge (n = 3 per infection group; n = 7 for mock-infected controls). C, Sperm counts at 4 and 7 days after Omicron or Delta virus challenge (n = 3 per infection group; n = 7 for controls). D, Serum concentration of testosterone and inhibin B in hamsters challenged with Delta or Omicron, at 4 or 7 dpi (n = 3 per infection group; n = 4 for controls). E, Viral loads in homogenized testicular tissue at 4 and 7 dpi were determined using quantitative reverse-transcription polymerase chain reaction. Data are presented as copies of SARS-CoV-2 RNA–dependent RNA polymerase (RdRp) gene per copy of β-actin in log scale (n = 3 per group). B–E, Data represent means with standard deviations. *P < .05; **P < .01 (Student's t-test). F, Representative HE-stained images of the testes and epididymis taken from hamsters intranasally infected with 10³ or 10⁵ PFU of (H1N1)pdm09 influenza virus. No histopathological changes were observed in testes or epididymis. Scale bars represent 500, 200, or 100 µm.

Serum levels of testosterone, but not inhibin B, were significantly reduced 7 days after Omicron or Delta infection (Figure 9D). Viral load was detected in some testes at 4 dpi (Figure 9E). However, intranasal challenge with 10³ or 10⁵ PFU of A(H1N1)pdm09 did not cause any histopathological changes in the testes at 4 dpi (Figure 9F), despite the finding of interstitial pneumonia in infected hamsters (Supplementary Figure 3).

Vaccination Protection From Testicular Damage After SARS-CoV-2 Challenge

Hamsters immunized with 2 doses of intramuscular inactivated whole virion vaccine 14 days apart [14] were intranasally challenged with 10³ PFU of SARS-CoV-2 HK-13 14 days after the second vaccination. Testes were examined at 4 and 28 dpi (Figure 10A). Serum neutralizing antibodies were detected from all vaccinated animals after challenge at 4 and 28 dpi, with geometric mean titers of 160 (standard error of mean [SEM] , 43.82) and 120 (SEM, 23.09), respectively. Testicles showed no histopathological changes in all 9 vaccinated hamsters (Figure 10B). One group of hamsters (n = 7) challenged with 10³ PFU at 3 days after the first vaccine dose showed no testicular histopathological damage at 4 dpi (Figure 10B), indicating that vaccination effectively protects testes from SARS-CoV-2 infection.

Figure 10.

Inactivated severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) vaccine can protect hamsters from testicular damage after intranasal challenged by SARS-CoV-2. A, Schema of experimental procedure. Groups of hamsters were immunized via intramuscular injection of inactivated SARS-CoV-2 whole-virion vaccine, in 2 doses given 14 days apart. The animals were intranasally challenged with 10³ plaque-forming unit (PFU) of SARS-CoV-2 14 days after the second vaccine dose or at 3 days after the first dose. Testicles were examined at 4 and 28 days after SARS-CoV-2 challenge (days post infection [dpi]). Hamsters receiving phosphate-buffered saline (PBS) injection served as unvaccinated controls. B, Summary of serum neutralizing antibody geometric mean titer (GMT) (with standard errors of mean) and representative hematoxylin-eosin (HE)–stained images of testes and epididymis at 4 and 28 days after SARS-CoV-2 challenge. Scale bar represents 200 µm. Abbreviation: ND, not done, .

DISCUSSION

Hamsters are now well established as a physiological small animal model for COVID-19 [12]. The animals recovered clinically by 28 dpi. In the current study, we showed that testicles were subclinically affected with acute decreases in sperm count and subsequent chronic asymmetric testicular atrophy associated with oligospermia and persistently low serum testosterone and inhibin B levels. These testes showed degeneration, necrosis, and inflammation at the interstitium containing Leydig cells, and also at the seminiferous tubules with disruption of orderly spermatogenesis from 4 to 120 dpi. Immunohistochemical staining of biomarkers showed that both Sertoli and germ cells of seminiferous tubules were severely affected. Viral N protein expression was occasionally found in germ, interstitial, and epididymal epithelial cells. These changes were generally more severe with higher virus inoculum. Moreover, intranasal challenge with wild-type SARS-CoV-2 or Omicron or Delta variants could consistently cause testicular damage. To prove the causative role of SARS-CoV-2 in the pathogenesis of testicular damage, we directly injected SARS-CoV-2 into the testicles transcutaneously, which demonstrated far more positive N protein expression in the interstitium and the presence of viral subgenomic RNA in testicular tissue than with the A(H1N1)pdm09 control virus.

Blood-borne and hepatitis viruses, including HIV and hepatitis B, C, and E virus, are associated with subfertility and decreased sperm count or quality [21]. Clinical cases of coxsackievirus-associated epididymo-orchitis have been reported [22], but only mumps virus is convincingly associated with human orchitis [23, 24]. However, no physiological animal models proving the causative role of these viruses in orchitis have been reported. Filovirus with a prolonged high level of viremia [25], leading to severe hemorrhagic fever can lead to persistent infection of immunoprivileged organs, including the central nervous system and the male reproductive tract in human and animal models [26]. Similarly, hepatitis E virus infection can be associated with a prolonged viremia and subfertility in human [27]. Some arboviruses, such as the Zika virus with prolonged viremia [28], is associated with sexual transmission, central nervous system infection, orchitis, and subfertility. Its causative role in epididymo-orchitis has been proved in an IFN-α/β receptor deficient mice model. In the current study we showed that SARS-CoV-2, with undetectable or transient viremia, could cause persistent testicular damage in the clinically relevant hamster model with self-limiting pneumonia.

Involvement of the male reproductive tract by SARS-CoV-2 has previously been suspected because angiotensin-converting enzyme 2 (ACE2), the receptor for SARS-CoV-2, and/or transmembrane protease serine 2 (TMPRSS2), enzyme for proteolytic activation of spike, are abundantly expressed in spermatogonia, Sertoli, Leydig, and myoid cells [29, 30]. Furthermore, case series of testicular or epididymal pain, histopathological changes of orchitis at postmortem examination, SARS-CoV-2 RNA detection in semen, reduced sperm counts, and decreased serum testosterone level have also been reported [31–34]. Besides conflicting clinical reports [8, 9], the testicular damage could be due to direct viral invasion or an indirect effect of systemic hypercytokinemia, endotheliitis, vasculitis, thrombosis, and hypoxia due to respiratory failure or shock in severe COVID-19 [2]. Two previous hamster studies demonstrated very low viral loads in testicles. However, one intranasal challenge study showed expression of testicular SARS-CoV-2 RNA without histopathological damage up to 1 month after infection, despite the detection of viral replication in ex vivo hamster testicular cells [35]. Another hamster study showed only oophoritis but not orchitis, despite patchy prostatitis and seminal vesiculitis [36].

Here, we demonstrated that increasing infectious SARS-CoV-2 dose was associated with worsening testicular damage, as evidenced by lower testicular weight at 120 dpi and more severe and persistent histopathological changes. This may explain the high incidence of testicular changes found in deceased patients [33, 37]. The higher virus inoculum may increase the likelihood of SARS-CoV-2 crossing the blood-testicular barrier into this immune-privileged organ. Although virus replication and leukocyte infiltration in testicles were far lower than in lungs, we speculate that acute damage at 4–7 dpi is more related to germ cells which are highly susceptible to the cytokines/chemokines produced by Sertoli cells and interstitial macrophages during their innate response against the small amount of SARS-CoV-2 going across the blood testicular barrier [38]. While the testicular cytokine/chemokine levels normalized between 42 and 120 dpi, the process of immune complex deposition and immunopathological damage might start after 7 dpi. As expected, the tissue mRNA expression level of proinflammatory cytokine/chemokines was associated with acute damage, and the apoptotic markers associated with both acute and chronic testicular damage. This is consistent with the pathogenesis of mumps orchitis, in which mumps virus triggers type 1 IFNs, TNF-α, IL-6, CXCL10, and MCP-1(monocyte chemoattractant protein 1), which induce apoptosis, disrupt the blood-testicular barrier, inhibit testosterone synthesis in Leydig cells, and recruit inflammatory cells into testes [24].

To prove the causative role of SARS-CoV-2 in the pathogenesis of testicular damage, we directly injected SARS-CoV-2 into animal testes to bypass the blood-testicular barrier. More abundant N protein expression was found in interstitial cells. Interstitial Leydig cells may be a target of SARS-CoV-2. Leydig cells are the primary source of testosterone in males. Its high concentration at seminiferous tubules is pivotal for maintaining testicular microvascular blood flow, Sertoli cell maturation, orderly spermatogenesis, and an intact blood-testicular barrier. Intratesticular inoculation of mumps [39] and Zika [40] virus were proposed in rodent models of orchitis. Finally, we showed that vaccination given as late as 3 days before virus challenge could protect animal testes against SARS-CoV-2–induced damage, similar to the protection in lungs [14].

Our findings here are limited to the hamster model we used. Few antibodies against hamster biomarkers for immunohistochemistry were available to locate infected cell types. We did not follow up beyond 120 dpi owing to limited availability of biosafety level 3 facilities. In summary, SARS-CoV-2 can cause acute and chronic testicular damage in hamsters and is consistent with the anecdotal reports of clinical orchitis and hypogonadism in men recovered from COVID-19. Long-term follow-up of sperm counts and sex hormone profiles in convalescent COVID-19 males is warranted.

Supplementary Data

Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Author Contributions. C. L., Z. Y., A. J. X. Z., J. F. W. C., and K. Y. Y. had roles in the study design, data collection, data analysis, data interpretation, and writing of the manuscript. W. S., F. L., Y. C., A. C. Y .L., Y. Z., B. H. Y. W., C. C. Y. Y., and J. P. C. conducted part of the experiments. M. Y.W. K., D. C. L., S. S., D. J., H. C., and K. K. W. T. had roles in the study design, experiments, data collection, and/or data analysis. All authors reviewed and approved the final version of the manuscript.

Disclaimer. The funding sources had no role in the study design, data collection, analysis, interpretation, or writing of the report.

Financial support. This work was supported by the Health and Medical Research Fund (grant COVID190121), the Food and Health Bureau, Government of the Hong Kong Special Administrative Region; the Theme-Based Research Scheme of Research Grants Council, Government of the Hong Kong Special Administrative Region (grant T11-709/21-N); Health@InnoHK, Innovation and Technology Commission, Government of the Hong Kong Special Administrative Region; the National Program on Key Research Project of China (grants 2020YFA0707500 and 2020YFA0707504); the National Natural Science Foundation of China Excellent Young Scientists Fund (Hong Kong and Macau) (grant 32122001); the Sanming Project of Medicine in Shenzhen, China (grant SZSM201911014); the High Level-Hospital Program, Health Commission of Guangdong Province, China; the research project of the Hainan Academician Innovation Platform (grant YSPTZX202004); the Emergency Key Program of Guangzhou Laboratory (grant EKPG22-01); the National Key Research and Development Programme on Public Security Risk Prevention and Control Emergency Project; the Consultancy Service for Enhancing Laboratory Surveillance of Emerging Infectious Diseases and Research Capability on Antimicrobial Resistance, Department of Health, Government of the Hong Kong Special Administrative Region; the University of Hong Kong Outstanding Young Researcher Award; the University of Hong Kong Li Ka Shing Faculty of Medicine Research Output Prize; and donations from the following: Michael Seak-Kan Tong, Richard Yu and Carol Yu, Shaw Foundation Hong Kong, May Tam Mak Mei Yin, the Providence Foundation (in memory of the late Lui Hac Minh), Lee Wan Keung Charity Foundation, Hong Kong Sanatorium & Hospital, Respiratory Viral Research Foundation, Hui Ming, Hui Hoy and Chow Sin Lan Charity Fund, the Chan Yin Chuen Memorial Charitable Foundation, Marina Man-Wai Lee, the Hong Kong Hainan Commercial Association South China Microbiology Research Fund, the Jessie & George Ho Charitable Foundation, Perfect Shape Medical, Kai Chong Tong, Tse Kam Ming Laurence, the Foo Oi Foundation, Betty Hing-Chu Lee, Ping Cham So, and the Lo Ying Shek Chi Wai Foundation.

Potential conflicts of interests. J. F. W. C. has received travel grants from Pfizer Hong Kong and Astellas Pharma Hong Kong and was an invited speaker for Gilead Sciences Hong Kong and Luminex. K. K. W. T. and K. Y. Y. report collaboration with Sinovac Biotech and China National Pharmaceutical Group (Sinopharm). All other authors report no potential conflicts.

All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

Author notes

C. L., Z. Y., A. J. X. Z., and J. F. W. C. contributed equally to this work.