All animal experiments were approved by the National Institute of Allergy and Infectious Diseases, Division of Intramural Research, Animal Care and Use Committee (ACUC) and adhered to National Institutes of Health (NIH) policies. The entire procedure to handle NHPs including injections or inoculations of virus, blood withdrawals, and biopsy and necropsy strictly followed NIH NHP regulations and guidelines. All animal handling and scientific procedures (e.g. virus injection, blood withdrawal, and biopsy) were performed under anesthesia with intramuscular injection of ketamine hydrochloride 10-25mg/kg. Investigators and animal care personnel provided additional comfort and care such as supplemental heat and IV, SQ, or oral fluids, consistent with the scientific integrity of the protocol. Animals that were not eating were tube-fed under sedation. Such animals were given metoclopramide 0.3 mg/kg IM 15 minutes prior to tube feeding to prevent vomiting.

Animals were euthanized in accordance with the 2007 Report of The American Veterinary Medical Association Panel on Euthanasia utilizing exsanguination via direct cardiac puncture subsequent to the induction of deep anesthesia by intravenous injection of 100 mg/kg of sodium pentobarbital (Fatal Plus®, Vortech Pharmaceuticals, Dearborn, MI). Death was ensured by administration of sodium pentobarbital.

MPXV Zaire 79 stocks were propagated in BSC-1 cells at a multiplicity of infection (MOI) of 0.1 for 3 days. Virus inoculum was crude virus prepared by 3 freeze/thawing cycles followed by centrifugation and ultrasonic treatment, then pelleted over a 36% sucrose cushion.

The experimental design is outlined in Table 1. Rhesus macaques (Macaca mulatta) of both sexes, ranging in weight from 4–8 kg, were housed in biosafety level 3 (BSL-3) bio-containment cages (Primate Products, Miami, FL). Such macaques were acclimated to a 12:12 hour light/dark cycle in a temperature and humidity controlled vivarium facility at NIH. All animals were acclimated to the study facility for a minimum of 2 weeks prior to the start of the study. Prior to enrollment, NHPs were screened and found to be seronegative for simian retrovirus, simian T cell leukemia, VACV, cowpox virus, and MPXV. Virus inoculations were performed by IV injection of desired virus doses in 1ml of phosphate buffered saline (PBS). At different time points before and after virus inoculation, blood was collected aseptically in ethylenediaminetetra-acetic acid (EDTA)-treated tubes. Blood samples were tested for changes in standard clinical serum chemistry values, virus load, and total cell counts of blood cell populations to monitor subject health. Clinical signs were evaluated daily. Biopsies of axillary LNs at day 5 and day 12 after virus inoculation in Experiment B and complete necropsies at the end of each experiment were performed following standard surgical procedures.

The data reported herein are results from two individual experiments. Experiment A included six MPXV-infected NHPs (2.5 x 10⁶ PFU Zaire 79 strain) and two normal controls. We performed the following analyses on samples from experiment A: 1), chemokine receptor staining from peripheral blood mononuclear cells (PBMCs) at different time points and LN cells from days 8 and 9 necropsy samples; 2), NK cell frequency and total NK cell number from whole blood and day8/9 LNs; and 3), NK cell functional analysis (CD107a, IFN-γ, TNF-α) from PBMCs and LN cell samples. In experiment B, five NHPs were infected with 1.5 x 10⁶ MPXV Zaire 79 strain and another two NHPs served as uninfected controls. We obtained the data for the total NK cell number in the blood and the Ki67+ expression data in the blood and LNs (days 5 and 12) from Experiment B.

The following antibodies (Abs) and conjugates were purchased from BD Pharmingen (San Jose, CA): CD3-Pacific Blue, -Alexa Fluor^® 700, -APC (SP34-2); CD8α-APC-H7 (SK1); CD8α-Pacific Blue (custom conjugate, SK1); CD14-Pacific Blue, -Alexa Fluor^® 488 (M5E2); CD16-APC, -PE (3G8); CD45-PerCP (D058-1283); CD56-PE-Cy7 (NCAM16.2); CD197 (CCR7)-Alexa Fluor^® 647; CD183 (CXCR3)-PerCP-Cy5.5 (1C6); CD195 (CCR5)-APC (3A9); CD196 (CCR6)-PerCP-Cy5.5 (11A9); human Ki67-PerCP-Cy5.5 (B56); CD107a-APC (H4A3); INFγ-PerCP-Cy5.5 (B27); TNFα-V450 (Mab11); mouse IgG2a-AlexaFluor^® 647, -APC (G155-178); and mouse IgG1-PerCP-Cy5.5 (MOPC-21). CD16-FITC (VEP13) was purchased from Miltenyi Biotech (Auburn, CA). CD159a (NKG2A)-PE (Z199) was obtained from Beckman Coulter (Miami, FL). Yellow LIVE/DEAD^® dye was purchased from Invitrogen (Carlsbad, CA).

EDTA blood was either treated with ammonium-chloride (ACK) lysis buffer to obtain leukocytes or layered over Ficoll density gradient to obtain PBMCs. To obtain leukocytes, 0.7 ml of whole blood was suspended in 3 mL of ACK buffer (Invitrogen, Carlsbad CA) for 2–3 min, centrifuged, and washed twice with PBS containing 2% fetal bovine serum (FBS). For the preparation of PBMCs, EDTA-blood was diluted 2-fold in PBS, layered over Ficoll solution (Sigma-Aldrich, St. Louis, MO), and centrifuged at 1000 x g at room temperature (RT) for 30 min. Then, the interface cell layer was collected and washed twice with PBS+2% FBS. The PBMCs were either directly used for flow cytometric staining or for NK cell functional assays.

To prepare cells from LNs, the axillary LN was placed in a cell strainer in a petri-dish containing 10 mL of PBS with 2% FBS. The tissues were gently ground with the tip of a syringe. The cells were collected and washed twice with PBS containing 2% FBS for further analysis.

To characterize NK cells and NK subsets, PBMCs or leukocytes (0.5–1 x 10⁶ cells) were incubated with a mixture of antibodies (Abs) to cell surface markers for 15–20 min at RT. For chemokine receptor staining, PBMCs were incubated at 37°C in RPMI-1640 medium containing 5% FBS for one hours followed by staining of cells with Abs against chemokine receptors at 37°C for one hour. Cells were then washed and subsequently stained with antibodies against CD3, CD8, CD16, CD56, and NKG2A for 30 min at RT. Then the cells were washed with PBS+2% FBS and were suspended in BD FACS lysing buffer (BD Biosciences, San Jose, CA) for 30–40 min to inactivate potential viruses before data acquisition. For intracellular staining, the cells were washed twice with PBS+2% FBS, then fixed with Cytofix/Cytoperm (BD Biosciences) for 30–40 min at RT. The cells were washed with BD Perm/Wash, followed by incubation with Abs to Ki-67 or cytokines. After washing with BD Perm/Wash, the cells were suspended in Perm/Wash buffer and were acquired using a LSR-Fortessa cytometer (Becton Dickinson, San Jose, CA). The data were analyzed with Flowjo software (TreeStar Inc., Ashland, OR) or with BD FACS Diva software (BD Biosciences).

The absolute number of NK cells in the blood was obtained in two steps. First, in BD Trucount^™ tubes (BD Biosciences), the absolute number of blood CD45+ leukocytes, CD3+, CD8+ and CD4+ T cells and B cells were obtained. In a second set of tubes, blood cells were stained with markers to distinguish NK cells and subsets including CD3, CD8, NKG2A, CD16, and CD56. The absolute number of CD16+, DN, CD56+, and DP subsets within the NKG2A gate was calculated based on the absolute number of CD3-CD8+ population from the Trucount tubes.

The absolute number of total live cells in each axillary LN was obtained by Trypan Blue exclusion, and the frequency of NKG2A+ NK cells within the lymphocyte gate was obtained by flow cytometric staining. Then the total NK cells (NKG2A+) and NK subset cell number were calculated based on the total LN cell number and the frequency of NK cells and NK subsets in the LNs.

To analyze NK cell function, we adapted a method published by Reeves et al. [31]. NK cells were enriched by depleting CD3+ T cells with CD3+ T cell microbeads (Miltenyi Biotech Auburn, CA). The resulting cell population contained less than 3% of CD3+ T cells. NK-enriched cells (1 x 10⁶ cells) were co-cultured with MHC-I negative human cell line 721.221 at a ratio of 5:1 for 10–12 h at 37°C in 5% CO₂ in serum-free medium, Aim-V^® (Life Technologies, Grand Island, NY). Cultures containing phorbol myristate acetate (50 ng/mL) and ionomycin (200 ng/mL) or medium alone served as positive and negative controls, respectively. Anti–CD107a-APC (clone H4A3) was added into the culture (1μL/mL) during the incubation period and BD Golgiplug and Golgistop (BD Biosciences) were added into the culture as manufacturer recommends. Following overnight culturing, the samples were stained using surface markers to reveal NK cells and NK subsets (CD3, CD8, NKG2A, CD16, and CD56). The cells were then fixed and permeabilized using BD Cytofix/cytoperm followed by intracellular cytokine staining with Abs against IFN-γ and TNF-α.

Viral load in whole blood was determined by quantitative PCR (qPCR) after DNA isolation from 50 µl of whole blood using Genfind v2 according to the manufacturer’s directions (Agencourt, Danvers, MA) as described previously [39]. Results were calculated by reference to the standard curve, multiplied by the appropriate dilution factor, and reported as gene copies per ml. The lower limit of detection of the synthetic oligo was 10 copies, while the upper limit was 1x10⁸ gene copies per ml. We set our limit of detection at 10,000 copies/ml blood so that we were well within the standard deviations that would account for dilution errors and extraction efficiencies.

Statistical and graphic analyses were completed using GraphPad Prism software (GraphPad Software Inc, San Diego, CA) and the R Project for statistical computing (R Core Team, Vienna Austria). Most experiments were tested using repeated measure two-way ANOVA with Bonferroni post-tests. Student t-tests were also used in some analysis. P values less than 0.05 (p<0.05) were assumed to be significant in all analyses. All analyses were verified using appropriate bootstrap and permutation re-sampling tests [40,41] in situations where the assumptions were not met for parametric T-tests and ANOVA or for analogous nonparametric Mann-Whitney and Kruskal-Wallis tests. Wherever needed, Bonferroni or Holm adjustments for multiple tests were used. Statistical analysis for the correlation between virus load and NK cell number was using both Pearson correlation and Spearman rank correlation test.

Previous studies with MPXV challenge in NHPs reported that MPXV infection induced lymphoid depletion and lymphadenopathy [3,4]. This raises the question whether MPXV infection induces changes in the number of lymphocytes including NK cells in NHPs. From blood samples collected at different time points (Table 1), the absolute numbers of lymphocytes and NKG2A+ NK cells or CD3-CD8+ NK-enriched cells were enumerated. The number of lymphocytes slightly decreased from baseline to day 2 and day 4 and then increased on average 3.0-fold at day 7 in experiment A (Figure S1A) and 2.1-fold at day 8 in experiment B (Figure S1B). After day 8, overall lymphocyte counts remained high in a range of 1.5-2-fold of the base line (Figure S1B). The baseline number of NK cells in rhesus macaques from our study was 116 cells/μl of blood (range from 26–232 cells/μl blood) (Figure 1A, Experiment A). At days 2 and 4 after MPXV inoculation, the absolute numbers of NK cells remained relatively unchanged. NK cell counts began to increase at around day 5, and peaked (average 2704 cells/μl blood) at day 7 after virus inoculation, representing an average increases of 23-fold over the baseline counts. At day 8, the NK cell number started to decline, but was still 17-fold higher than the baseline (Figure 1A). Accordingly, the frequency of NK cells of the total lymphocyte population rose from an average 4.7±2.3% at baseline to 41.4±7.0% at day 7 and 36.0±5.6% at day 8 (Figure 1B, Experiment A). Results from experiment B confirmed the kinetics of NK cell numbers with a peak at day 8, representing an average increase from 5.7±2.6% of the total lymphocyte population at baseline to 35.4 ±8.2% at day 8, followed by a decline close to baseline at day 21 (Figure 1C, 1D). The changes in NK cell numbers upon MPXV infection varied markedly among individual NHPs. For example, NHP AT25S showed a 48-fold increase (4969 cells/μl blood at day 7 versus 104 cells/μl blood at baseline), while NHP iD3 had only a 10-fold increase (1627 versus157 NK cells/μl blood at day 7 or day 0, respectively) (Figure 1A).

We questioned if all NK subsets or specific subset(s) increased in the blood following MPXV challenge. NK cell subsets were distinguished based on CD16 and CD56 expression within the NKG2A+ NK cell gate (Figure 1E) [31]. The frequency of CD16+ NK subset within the total NK cell population decreased from 78.7 ± 4.5% at day 0 to 28.8 ± 21.7% (p< 0.0001 ) at day 7 post MPXV inoculation (Figure 1F). The frequency of DN cells increased from 4.6 ± 2.7% to 41.4 ± 16.7% (p<0.0001), becoming the dominant NK cell subset in the blood. In addition, CD56+ NK cells also increased from 5.5 ± 2.4% to 19.4 ± 9.7% (p=0.18) and DP NK cell frequency remained unchanged (10.7 ± 1.2% and 10.9 ± 8.5%, p>0.05). The absolute number of all NK subsets at day 7 postinoculation significantly increased (p<0.001 or p<0.0001) upon MPXV infection (Figure 1G). Among them, the DN population showed maximal increase (around 100-fold). For CD16+, CD56+, and DP NK cells, the increases were 8.1-, 71-, and 30-fold on average (Figure 1G).

We questioned if MPXV infection induced changes in NK cell number and composition in the LNs. Our analysis showed that the frequency of total NK cells among lymphocytes increased about 8.4-fold from an average of 0.55% in control NHPs to 4.6% in MPXV-infected NHPs (Figure 2A). Accordingly, the total number of NK cells per axillary LN dramatically increased from an average of 1.1x10⁶ cells in control NHPs to 50.7 x10⁶ cells in MPXV-infected NHPs (Figure 2B). The magnitude of change in NK cell numbers varied greatly among individual MPXV-infected NHPs. For example, the total NK cell number was 14.1 x10⁶ per LN in one NHP and it was 147.5 x10⁶ in another NHP (Figure 2B).

Analysis of NK cell subsets in the axillary LNs from control NHPs confirmed previous reports [31,42] that DN and CD56+ cells were the major NK cell subpopulations in the LNs of uninfected NHPs (Figure 2A, B and C). At days 8–9 post-inoculation, the CD16+ and DP NK cells in the LNs increased at variable frequencies. For example in NHP iD3 (Figure 2C), CD16+ population reached 29.3% of total NK cells and DP cells accounted for 15% of total NK cells. However, in NHP DE3V, the CD16+ and DP populations accounted for 5.26% and 42.3% respectively, of the total NK cells. Across all MPXV-infected NHPs, the change in CD16+ cell frequency was not significant (p>0.05). DN cell frequency was reduced from average of 50.6% in control NHPs to 20.3% in MPXV-infected NHPs (p<0.01). CD56+ cell frequency remained unchanged (p>0.05). Regardless the increase or decrease in cell frequency, the absolute number of all subsets significantly increased as a result of the marked increase of total NK cell number in the LNs (p<0.01, p<0.001, or p<0.0001) (Figure 2B). Of the four NK cell subsets, DP cells on average displayed the greatest magnitude of increase. This increase in DP cells is partially a consequence of the extremely high number of total NK cells (147.5 x 10⁶) and DP cells (60.1 x 10⁶, 42.3% of total NK cells) in NHP DE3V.

In comparing the NK cell increases in the blood and the LNs, the magnitude of NK cell increases in the blood is not similarly observed in the LNs for a given NHP. For example, NHP AT25S in the blood had the highest NK cell number in the group (Figure 1A), but the LN NK cell number was relatively low compared to other NHPs (Figure S2). In contrast, NHP DE3V had the highest NK cell number in the LN, but the NK cell number in the blood in this NHP is relatively low (Figure S2 and Figure 1A).

The marked increase of NK cells in the blood and LNs could be largely due to cell proliferation. In a previous study in mice infected with either MCMV or VACV, peripheral NK cells rapidly proliferated and peaked at day 6 postinoculation [43]. Similarly, ECTV infection also induced NK proliferation in mice [20]. To assess NK cell proliferation in our study, we examined Ki67 expression in the NK cells from the blood and LNs. NK cells from the blood of normal control NHPs and NHPs before virus inoculation showed a low frequency of Ki67 staining (<5%), and the Ki67 expression frequency at day 2 after virus inoculation was similar to background (Figure 3A). However, at days 5–8 after MPXV inoculation, about 30% of NK cells in the blood expressed Ki67 protein (Figure 3A), a time frame consistent with previous reports [20,43] and with the NK cell number change in the blood (Figure 1A). The frequency of Ki67+ cells in the blood gradually reduced to baseline at days 21–30. When we examined the proliferation of NK cell subsets, CD16-negative cells in the blood including DN cells and CD56+ cells dominantly expressed Ki67 protein (Figure 3B). While a low frequency of CD16+ (~15-20%) and DP (~5-8%) populations also expressed Ki67, the Ki67+ cells in these two populations tended to express a low intensity of CD16 protein (data not shown).

We next measured Ki67 expression in NK cells in LNs at days 5 and 12 postinoculation and found that a higher proportion of LN NK cells from the MPXV-infected NHPs at day 5 expressed Ki67 (average 26%) compared to NK cells from control NHPs (average 7.9 %) (Figure 3C). In addition, day 5 LNs from infected NHPs seemed to have a higher frequency of Ki67+ cells than day 12, supporting the notion that cell proliferation peaked at an early stage of virus infection (Figure 3C, 3D) [20,43]. Similar to blood NK cells, LN CD16^- NK cells (DN and CD56+ NK cells) displayed higher Ki67 staining than CD16⁺ (CD16+ and DP) cells (Figure 3D).

To study the potential of NK cells to migrate to infected peripheral tissues upon MPXV infection, we assessed the expression of chemokine receptors CXCR3, CCR5, CCR6, and CCR7 on each NK cell subset. Similar to result from others [31,44], our analysis of NK cells from pre-inoculation blood showed that CD56+ cells (average 40%) and DN cells ( average 42%) were the major subsets expressing CCR5 (Figures 4A) . In contrast, only a very small proportion of CD16+ and DP NK cells express a low intensity of CCR5. The frequency of CCR5-expressing cells in all subsets remained unchanged at day 2 after MPXV infection and unchanged for CD16+, DN, and DP NK cells at day 4, but moderately increased in CD56+ cells (from average 40 to 60%, p>0.05) at day 4. At day 7, however the average frequency of CCR5-expressing cells in all subsets largely increased but did not attain a statistical significance (Figure 4A). In addition, the intensity of CCR5 expression also increased at day 7 after MPXV inoculation as shown in a representative NHP in Figure 4B. Interestingly, the CCR5-expressing cells in CD16+ and DP population were CD16^dim cells (low intensity of CD16 expression), suggesting the CD16^dim population might be preferentially recruited to tissues compared to CD16^high NK cells (Figure 4B).

CCR7 is the chemokine receptor required for T and NK cells to home to LNs [45-47]. Similar to result from other reports [31,46], our results showed that only CD56+ NK cells at steady state expressed CCR7 (Figure 4A). Following MPXV infection, the frequency of CCR7+ CD56+ cells gradually decreased from 78% at day 0 to 50–60% on days 2–4 and to a low or negative frequency at days 7–8. MPXV infection did not induce a distinguishable change on CCR7 expression in other three NK populations that were negative at steady state and during MPXV infection (Figure 4A).

CXCR3 has been shown to mediate the migration of effector T cells and NK cells to lymphoid tissues and/or inflamed tissues [48,49]. Consistent with reports from others, we found that about 90% of circulating CD56+ NK cells and 40% of DN cells expressed CXCR3 at steady state, but CD16+ and DP cells did not express CXCR3 (Figure 4A). Following MPXV inoculation, the frequency of CD56+ cells and DN cells that expressed CXCR3 was first reduced (p<0.05 CD56+, p>0.05 DN) at days 2 and 4, and then largely recovered at days 7–8 (p<0.05 for DN cells). In addition, about 20-25% of CD16+ cells and 60% of DP NK cells also expressed CXCR3at days 7–8.

The role of CCR6 in cell migration is less clear than the other chemokine receptors. Some investigators indicated that CCR6 may play a role in recruiting NK cells to tissues [31]. Also consistent with a previous report of SIV-infected macaques [31], results from our data showed that various proportions of NK subsets before virus inoculation express CCR6. MPXV infection ablated CCR6 expression on all four NK cell subsets (Figure 4A).

We further analyzed chemokine receptor expression on NK cells in the axillary LNs from control NHPs and from NHPs at necropsy days 8–9 following MPXV inoculation. In control NHPs, about 45–80% of all four NK subsets from LNs expressed CCR5, CCR6, and CXCR3, and only CD56+ and DP cells expressed CCR7 (Figure 5). Upon MPXV infection, CCR5-expressing NK cells greatly increased (p<0.01) and almost all LN NK cells expressed CCR5 (Figure 5). LN CD16+ and DN NK cells did not express CCR7 at steady state and during MPXV infection. However, the CCR7-expressing cells in CD56+ and DP population were markedly reduced (p<0.01) upon MPXV infection, a result consistent with that from blood NK cells. The expression of CXCR3 did not change in CD16+ and DN cells during MPXV infection, but slightly decreased in CD56+ and DP cells (p>0.05). Similar to circulating NK cells, almost all NK cells in the LN lost CCR6 expression following MPXV infection.

We next asked whether MPXV infection induces any changes in NK cell functions, including cytotoxicity capacity and secretion of cytokines. We assessed the expression of CD107a, a molecular marker for degranulation and cytotoxicity [50,51], and intracellular expression of IFN-γ and TNF-α. Enriched NK cells from blood and LNs were stimulated with MHC-I-devoid cell line 721.221, and the expression of surface CD107a and intracellular cytokines were measured. Consistent with published reports [31], circulating CD16⁺ NK cells from control NHPs expressed CD107a and barely expressed IFN-γ and TNF-α (Figure 6), supporting their major role in killing infected cells. DN cells, on the other hand, displayed strong expression for both CD107a and cytokines. It is believed that CD56+ NK cells are specialized to produce cytokines, such as IFN-γ and TNF-α [31,33] and our data from normal LN control NK cell cultures support the cytokine production role of CD56+ NK cells by showing that about 20% of CD56+ cells stained positive for IFN-γ and TNF-α (Figure 6). However, we also observed that 20% of CD56+ NK cells from the normal LN culture expressed CD107a. Similar to the NK cells in the blood, DN cells from the normal control LNs expressed both CD107a and cytokines. Upon MPXV infection, the frequency of CD107a-expressing CD16+ cells from the blood and the CD56+ cells from the LN remained unchanged. However, CD107a expression in the DN cells from both blood and LNs was significantly reduced (p<0.01). Most strikingly, almost all the NK cells, regardless of the source (blood and LNs) and subset, lost their capacity to express both IFN-γ and TNF-α, indicating a crippled NK cell cytokine response following MPXV infection.

NHPs in both experiments A and B received a sub-lethal dose of MPXV and all NHPs survived through each individual study end point (Table 1). In general, all NHPs developed typical clinical signs of MPXV infection including fever, weight loss and skin lesions [4]. No apparent difference was observed regarding the clinical measurements (CBC/diff and serum chemistry) and clinical signs among the NHPs from both experiments. No clear correlation between the maximal lesion number and the peak NK cell number at day 7-8 was established (R=0.7 and p=0.233, data not shown). Virus load (viral DNA copy number measured by qPCR) was used in previous studies as a parameter to predict disease progression [52,53]. Our studies using qPCR test of whole blood samples showed the onset of positive MPXV DNA was from days 2-7, roughly paralleling the increases in NK cell number and lymphocyte number in the blood. Viremia lasted from a few days to 10-11 days (data not shown). The peak MPXV genome copies varied among individual NHPs and between the two experiments (Table 2). One out of six NHPs from experiment A and two out of four NHPs from experiment B were MPXV negative at all time-points. The sample from NHP DC98 was contaminated and thus not included in analysis. Statistical analysis showed no clear correlation between virus load and NK cell number (p>0.05, data not shown).