ABSTRACT
The process of vaccination introduced by Jenner generated immunity against smallpox and ultimately led to the eradication of the disease. Procedurally, in modern times, the virus is introduced into patients via a process called scarification, performed with a bifurcated needle containing a small amount of virus. What was unappreciated was the role that scarification itself plays in generating protective immunity. In rabbits, protection from lethal disease is induced by intradermal injection of vaccinia virus, whereas a protective response occurs within the first 2 min after scarification with or without virus, suggesting that the scarification process itself is a major contributor to immunoprotection.
IMPORTANCE These results show the importance of local nonspecific immunity in controlling poxvirus infections and indicate that the process of scarification should be critically considered during the development of vaccination protocols for other infectious agents.
INTRODUCTION
Vaccinia virus (VV) has historically been used as the smallpox vaccine (1). Vaccinia virus has been intensively studied at the molecular level over the years and has served as a surrogate virus for the development of antipoxvirus drugs (2, 3) and for vaccine development (4). Animal models are a critical part of the in vivo evaluation of any smallpox vaccine as well as the disease process itself. Mouse (5, 6), nonhuman primate (7, 8), rabbit (9), and prairie dog (10) animal models have been developed. While all the models have their own strengths and weaknesses, rabbits infected intradermally (i.d.) have proven to be one of the most sensitive models in which to study orthopoxvirus infections, including VV. i.d. inoculation of VV has the advantage that very low doses of virus (<100 PFU) cause lethal disease within 8 days following infection, generates a disseminated infection, and demonstrates facile animal-to-animal spread (9).
Interestingly, earlier investigations had reported little to no mortality associated with the inoculation of New Zealand White (NZW) rabbits with VV (11, 12). Indeed, the same rabbits have also been utilized for decades for the production of polyclonal antibodies against vaccinia virus and vaccinia virus recombinants. However, upon closer investigation of the literature, it was found that most laboratories routinely inoculated their rabbits via scarification or multiple injections (in essence, scarification) due to the ease of the procedure and the generation of high yields of anti-VV antisera (13).
It is relevant to briefly review some aspects of the smallpox eradication program which resulted in the elimination of natural smallpox and is one of the greatest success stories in the history of public health (14). The eradiation campaign relied heavily on the use of the bifurcated needle to vaccinate individuals and allowed otherwise medically untrained personnel to administer the vaccine. It is important to recognize, however, that administration of the vaccine via the bifurcated needle is a reproducible method of scarification which introduces a relatively fixed small volume of a high-titer virus stock (∼1 × 108/ml) into the skin to generate immunoprotection with few if any adverse effects or signs of disease in the vast majority of vaccinated humans or animals.
However, we have shown in our laboratory that i.d. infection of rabbits with either VV or rabbitpox virus (RPV; a VV variant known to be particularly lethal in rabbits) caused a lethal disease within 8 days following infection, yet it is the accepted practice to scarify rabbits with infectious vaccinia virus to raise anti-vaccinia virus antibodies in rabbits. Given these seemingly divergent results, the influence of the route of inoculation was more carefully examined. We have determined that lethal disease as opposed to a protective response in rabbits depends on how the virus is inoculated (intradermal versus scarification) and that scarification mediates a protection that is independent of the virus and that occurs within the first 2 min. Surprisingly, therefore, protection against lethal intradermal infections is generated through the physical act of scarification whether virus is present or not. We propose that scarification produces a local skin injury which triggers protective elements of a nonspecific but strong innate protective immune response which confers protection against i.d. VV infection.
MATERIALS AND METHODS
Cells and viruses. (i) Cell culture.
CV-1 cells were maintained in minimum essential medium (MEM) with Earle's salts (Gibco, Grand Island, NY) supplemented with 2 mM glutamine (Media Tech, Herndon, VA), 50 U/ml penicillin G and 50 μg/ml streptomycin (Media Tech), 1 mM sodium pyruvate (Media Tech), 0.1 mM nonessential amino acids (Media Tech), and 10% (vol/vol) fetal bovine serum (Gibco).
(ii) Virus growth.
VV WR was kindly provided by Richard Condit, University of Florida. VV was grown and titers were determined on CV-1 cells using standard methods. Virus was pad purified over 36% sucrose using standard methods and resuspended in phosphate-buffered saline (PBS) (15, 16).
Animals and animal methodology. (i) Animal husbandry.
Nine-week-old (3- to 4-lb) female New Zealand White rabbits were obtained from Myrtle's Rabbitry (Thompsons Station, TN). The animals were housed in standard single cages at 20°C on a 12-h light/12-h dark regime and fed standard rabbit pellets ad libitum.
(ii) Infection by intradermal inoculation.
Infection was performed by shaving of both thighs of the rabbit and sterilization with an isopropanol wipe. VV (1,000 PFU) was diluted in 0.2 ml PBS, and 0.1 ml was injected bilaterally intradermally in the middle of the shaved area using a 27-gauge needle, delivering 500 PFU per flank and 1,000 PFU total per animal. Mock-infected rabbits were injected with PBS alone. We have never observed any clinical symptoms in any PBS-infected animals. Therefore, the data from i.d. infected and scarified animals from an earlier publication were used as the data for control animals for this study in an effort to conserve the number of animals used for these experiments (9).
(iii) Infection by scarification.
Infection was performed by shaving of both thighs of the rabbit and sterilization with an isopropanol wipe. A bifurcated needle was dipped into the virus solution (concentration, 109 PFU/ml VV diluted in PBS), the presence of liquid between the prongs was visually confirmed, and the bifurcated needle was jabbed into the skin of the rabbit 15 to 20 times until a drop of blood was observed. The jabbing was precise, giving a site of less than 1.5 cm in diameter. The animals were scarified bilaterally. Mock-infected rabbits were scarified with PBS.
(iv) Ring scarification.
Rabbits were shaved and flanks were prepared as described above. Animals were either infected by scarification followed by i.d. inoculation or vice versa. Animals in the i.d. inoculation and then scarification group received the intradermal infection (500 PFU bilaterally for a total dose of 1,000 PFU VV), followed immediately by circular (3 cm) scarification around the bubble created by the intradermal injection. Animals in the scarification and then i.d. inoculation group received scarification in a circle of approximately 3 cm in diameter, followed immediately by the intradermal injection of 500 PFU VV within the ringed area of scarification. Each flank received the dual infection prior to infection of the second flank of the animal.
(v) Monitoring of animals.
A microchip was placed in each rabbit at the time of infection to transmit body temperature and identification number to a DAS-5007 reader (Bio Medic Data Systems, Seaford, DE). Weight, temperature, respiration and heart rates, and physical observations of the rabbits (grooming habits, facial swelling, secretions, condition of primary lesion, presence of secondary lesions, and respiratory distress) were recorded daily. Criteria for euthanizing the rabbits included open-mouth breathing, severe difficulty in breathing, or a loss of body weight greater than 20% from the beginning initial weight. These criteria are the established euthanasia criteria for the rabbit model previously published and correspond to mortality; however, no animals were permitted to succumb to disease (9, 17, 18). In all animals noted in this publication, euthanasia was dictated by severe respiratory failure, defined as less than 40 breaths per minute and/or open-mouth breathing upon stimulation. All animal procedures were carried out according to the University of Florida IACUC guidelines.
Clinical scores were determined daily as a numerical representation of disease for each animal. The clinical score included weight change, body temperature, heart and respiration rates, as well as the condition of the primary and secondary lesions and ranged from 0 to 34, with any animal with a clinical score over 20 considered exhibiting advanced or severe disease. These scoring parameters are presented in Fig. 1 and allow quantification of disease, including both quantitative and qualitative, often subtle signs of disease. These clinical scoring criteria have been successfully used in multiple studies to quantify disease in rabbits infected with both RPV and VV.
FIG 1.
Clinical parameters and scores for rabbits infected with vaccinia virus and rabbitpox virus. Each clinical parameter was measured daily and scored according to the outlined scoring system, and the resulting scores were then combined to generate the clinical index, a measure of disease severity.
Statistics.
Survival analysis was performed via Kaplan-Meier survival analysis using the SigmaStat program. The t test was performed in Microsoft Excel software.
Tissue processing. (i) Quantitative analysis of primary lesions.
Briefly, the site of infection (primary lesions) was isolated, excised, and sectioned in quarters horizontally so that each section contained parts of the middle and outer border of the lesion from the epidermis through the dermis. The sample was analyzed, together with tissue samples, as described below.
(ii) Quantitative real-time quantification of virus from tissues.
Two sections of tissue from euthanized animals were flash frozen until processing for determination of the virus titer. Tissue sections were processed for either traditional plaque assay or quantitative real-time PCR (qRT-PCR). All samples were assayed for virus quantification via quantitative PCR, with the results for a subset of the samples being verified by traditional plaque assay. All data are presented as the number of PFU equivalents, as the standards for the PCR were derived from the concentration consisting of the number of PFU of virus. Sections of each tissue were removed in such a fashion as to obtain a representative organ tissue sample from the same general anatomical site of each organ. Primary lesions were excised from the flank in the entirety and sectioned into quarters, with each section containing the necrotic center (when present) through the outer perimeter of the lesion.
Samples for the traditional plaque assay were processed as follows: tissue specimens were homogenized in 2-ml screw-top tubes (Sarstedt) with ∼1.0 ml sterile 24-grit silicon carbide (Electro Abrasives, Buffalo, NY) and 1 ml growth medium without serum (minimum essential medium; Invitrogen) using a Mini-Beadbeater 8 (Biospec, Bartlesville, OK) for 1.5 min on the homogenization setting. Virus titers in the supernatant were then evaluated by virus plaque formation on CV-1 cells using standard procedures (15, 16).
qRT-PCR was performed using the VV RPO22 gene as the target for analysis. DNA for qRT-PCR was isolated from 10 mg of tissue samples using a DNeasy kit following the manufacturer's recommendations (Qiagen). The DNA was then utilized to estimate the titer of VV via qRT-PCR using an iCycler system (Bio-Rad). A standard curve of purified virus was used in the development of the assay, with no differences being observed between DNA isolated from purified virus and DNA isolated from uninfected tissue spiked with known amounts of purified virus. Samples processed for conventional plaque assays or qRT-PCR provided comparable results and showed no statistically significant differences (data not shown).
Primers and probes were manufactured using the ABI Assays-by-Design custom TaqMan gene expression assay service. The primers used were 5′-TGTGTTTCAGGTATTTAACGAATCATCCA-3′ and 5′-CTTGTTATGACCTTTTTGAAGATACGATGTT-3′. The probe used was FAM-CCGGTTGATGATGATTATG-NFQ, where FAM is 6-carboxyfluorescein. qRT-PCR was performed using TaqMan Universal PCR master mix and No AmpErase uracil-N-glycosylase (ABI) following the manufacturer's directions.
(iii) Histopathology.
Tissue sections for histopathology were fixed in 10% buffered formalin (Fisher Scientific) for 12 to 18 h at room temperature and paraffin embedded. Processing and sectioning were performed by the CTAC Histology Core at the McKnight Brain Institute at the University of Florida. Sections of tissue (5 μm) were stained with hematoxylin-eosin using standard protocols.
Microarrays.
Sections of primary lesions were flash frozen in liquid nitrogen and maintained at −80°C until they were processed for RNA isolation. Primary lesions were homogenized manually with a mortar and pestle in the presence of lysis buffer (RNeasy kit; Qiagen) and 24-grit silicon carbide (Electro Abrasives, Buffalo, NY). RNA isolation was performed with the RNeasy kit (Qiagen) following the manufacturer's protocol. Normal skin was isolated from uninfected animals and treated as described above for the sections of primary lesions.
Processing of the RNA for microarrays was performed by the University of Florida's Interdisciplinary Center for Biotechnology Research Gene Expression Core Facility (Gainesville, FL). Total purified RNA (2 μg) was used for generation of cRNA. The cRNA served as the template for in vitro transcription, during which one of two cyanine-labeled nucleotides (PerkinElmer, Wellesley, MA) was incorporated into the synthesized cRNA. The two-dye system was used for the Agilent microarrays, with normal skin being used as the reference. Fifteen cyanine 3-cytosine triphosphate (Cy3) labeling reactions were performed on the normal skin, and the reaction mixtures were pooled and used for each microarray to limit the variability of the reference. Primary lesion samples were labeled with cyanine 5-cytosine triphosphate (Cy5). The in vitro transcription reaction mixtures were cleaned with RNeasy minicolumns (Qiagen, Valencia, CA). The samples were then fragmented, hybridized, and detected on a rabbit-specific custom array (Custom Microarray GE 8×15K; Agilent Technologies) (19).
The analysis of the microarray data was performed using the BRB Array Tools (version 4.2.0 Beta_1_p1) program (NIH). The median probe intensity was used to normalize the arrays. Probe sets were not analyzed if the number missing exceeded 20%. The entire set of arrays (those for 4, 24, and 48 h postinfection for intradermal injection and scarification) was subjected to an F test with random variance and a significance level of 0.01, and all arrays were assigned to classes on the basis of the treatment and time (hour) postinfection that the lesions were harvested. The analysis was performed with all the arrays included and then by time point to analyze the differences occurring at each time point.
Microarray data accession numbers.
The data discussed in this publication have been deposited in NCBI's Gene Expression Omnibus (GEO) database and are accessible through GEO Series accession number GSE57564.
RESULTS
Intradermal versus scarification infection of rabbits.
The observation that intradermal (i.d.) injection of both vaccinia virus (VV) and rabbitpox virus (RPV) is lethal in New Zealand White (NZW) rabbits has been previously described by our laboratory, in which the established the 50% lethal dose of VV was 10 PFU (9). Here we have more carefully evaluated the role of various routes of infection of VV in NZW rabbits in the outcome of disease. While our laboratory routinely uses 100 to 1,000 PFU in 100 μl of solution for i.d. infection, a bifurcated needle administers about 10 μl of solution. Several escalating concentrations of VV beginning with 1,000 PFU/ml were used, and a few signs of disease were observed with doses up to 1 × 107 PFU/ml (data not shown). Therefore, for the experiments reported here, virus at 109 PFU/ml was used for all scarification experiments with the bifurcated needle.
As previously observed, rabbits infected i.d. with 1,000 PFU VV demonstrated 80% mortality by 8 days postinfection (dpi) (9). Conversely, animals infected with VV via scarification using a bifurcated needed dipped into a solution with 109 PFU/ml VV demonstrated a relatively low mortality rate of only 15% by 8 dpi (Fig. 2A). All scarified animals survived infection at lower doses. The few animals infected via scarification that were euthanized due to severe disease had a disease course very similar to that of i.d. infected animals. In Fig. 2, while the number of animals is small, the trends for all measures are clear and in our experience accurately represent what is observed in the laboratory.
FIG 2.
Disease characteristics of animals infected with VV either via intradermal injection of 1,000 PFU or by scarification with a solution of 109 PFU/ml compared to those of uninfected (PBS-treated) animals. Data for PBS-treated animals were from a historical collection of data for animals intradermally injected with PBS. Animals were monitored daily for clinical signs of disease. (A) Survival (P = 0.096) for i.d. infected animals versus controls; (B) weight change over time; (C) average body temperature over time; (D) average clinical score over time. Error bars represent SEMs. Three rabbits received PBS, five rabbits were infected i.d., and eight rabbits were scarified. ⧫, PBS; ●, i.d. inoculation; ■, scarification. P was 0.02 for both temperatures and clinical scores at 6 dpi for the difference between animals infected i.d. and by scarification; 6 dpi was the final day on which all animals were alive.
Infected animals underwent complete physical examinations daily that allowed multiple parameters to be compared and assembled into a quantitative measure of clinical disease severity (the clinical index) (Fig. 1). These parameters included weight change, body temperature, and clinical score. As would be expected, there were large differences between the two groups in many aspects of disease severity, with rabbits infected via scarification presenting as a group an overall less severe disease phenotype. For example, scarified animals exhibited relatively little weight loss compared to i.d. infected animals, which exhibited a maximum weight loss of approximately 8% from their beginning weight. While 8% is not significant, in juvenile animals that are expected to gain weight daily, this weight loss is in reality a difference in weight of over 20% compared to the weight of uninfected age-matched control animals (Fig. 2B).
Body temperature has been observed to be an excellent indicator of both disease and the host response to infection (9, 17, 18, 20). Therefore, the timing of the change in body temperature, as well as the maximum, is important in gauging the level of response to an infection. Given that the scarified animals survived the infection, it was not surprising that these animals exhibited a fever, as defined by a body temperature of over 39.5°C, measured by the subcutaneous temperature transponders, 24 h before the i.d. infected animals. Scarified animals exhibited a fever beginning at 2 dpi, while i.d. infected animals did not exhibit a fever until 3 dpi. The time of onset of fever differed between scarified animals that survived and the few that exhibited severe disease and were euthanized, where scarified animals that succumbed to lethal disease exhibited body temperature profiles similar to those of i.d. infected animals (Fig. 2C).
The average clinical scores over time give a quantitative representation of the severity of disease. For the scarified animals that survived the infection, a maximum score of 10.5 was observed. This score stemmed from no weight gain, elevated body temperature, necrosis at the primary site of infection, and the presence of secondary lesions. The maximal clinical score for scarified animals that were euthanized for severe disease was 24.5 and stemmed from severe respiratory distress, elevated body temperature, some weight loss, necrosis at the primary lesion, numerous secondary lesions, and a generalized depression. In comparison, the i.d. infected animals had a maximum clinical score of 27.5, which stemmed from severe respiratory distress, elevated body temperature, major weight loss, necrosis at the primary lesion, numerous secondary lesions, and a generalized depression (Fig. 2D).
The development of the primary lesion for both scarified and i.d. infected animals was initially very similar. There was a local reaction that began with swelling and redness at the site of infection at 1 dpi and that by 4 dpi progressed to necrosis with a thickening of the epidermis surrounding the primary lesion (Fig. 3). Lesion development was far more extensive in i.d. infected animals. The appearance of secondary lesions, an indicator of viremia for both groups, occurred at 3 to 4 dpi, although the scarified animals typically had more numerous and larger secondary lesions on the ears. This may be in part due to the fact that the scarified animals survived longer, and therefore, the secondary lesions showed more significant development.
FIG 3.
Disease progression over time after infection with VV i.d. and by scarification. Representative photographs of ears (showing secondary lesions) and flanks (showing primary lesions) for VV-infected animals. Results at 3 dpi (A and D), 5 dpi (B and E), and 7 dpi (C and F) for i.d. infected animals and at 3 dpi (G and J), 5 dpi (H and K), and 7 dpi (I and L) for animals infected via scarification are shown. Bars, 1 cm.
Virus dissemination.
The ability of the virus to disseminate when administered via scarification compared to that when administered i.d. was determined at 3 and 7 dpi, representing early and late times of infection, respectively (Table 1), as measured by determination of the number of PFU equivalents from quantitative PCR measurements. There was no difference in the ability of the virus to replicate at the primary site of infection (approximately 104 PFU equivalents/mg of tissue). Although there was no difference in total titer, it is important to point out that the overall lesion size was vastly different between the two groups at late times postinfection, with the average size of a lesion after i.d. infection being 10 cm in diameter and that of a lesion after scarification being 3.5 cm in diameter by 7 dpi. Therefore, the total virus burden in the lesion induced by scarification was at least 2 orders of magnitude lower. Virus was capable of spreading to distal sites, as shown in Table 1 for virus spread at 3 dpi, where the levels of virus in the spleen, gonad, lung, and popliteal lymph nodes in the two sets of animals generally showed no significant differences. The exception was the liver, where no virus was detected at 3 dpi in the scarification group but virus was present at 104 PFU/g in the i.d. infected group at 3 dpi. This difference in liver titers was not present late in infection (7 dpi), when the scarified animals clinically showed signs of recovering, and may be an indicator of a slower progression of disease. At 7 dpi, the levels of virus in the spleen of scarified animals were nearly 100-fold lower than those in the spleen of i.d. infected animals. By 7 dpi, the scarified animals had cleared the virus from the gonads, while the i.d. infected animals maintained a significant viral load (104 PFU/g tissue). Titers determined late in infection (10 dpi) for the scarified animals showed a clearance of the virus from the internal organs, with the virus remaining only at the primary lesion, which is theorized to be contributed by the scab on the healing lesion (data not shown).
TABLE 1.
Virus titers in tissues harvested from rabbits infected via either i.d. injection or scarification at 3 and 7 dpi
| Tissue | Titera (log10 PFU equivalent/mg tissue) |
|||
|---|---|---|---|---|
| 3 dpi |
7 dpi |
|||
| Intradermal injection | Scarification | Intradermal injection | Scarification | |
| Primary lesion | 3.9 (8/8) | 4.0 (8/8) | 4.7 (12/12) | 4.4 (8/8) |
| Popliteal lymph node | 4.3 (8/8) | 4.4 (5/8) | 4.1 (5/10) | 4.0 (2/7) |
| Spleen | 3.5 (4/4) | 3.4 (2/4) | 4.8 (2/6) | 2.6 (1/4) |
| Liver | 3.3 (2/3) | <1.1 (0/3) | 4.3 (3/6) | 4.1 (1/3) |
| Gonad | 3.9 (3/3) | 3.9 (1/4) | 4.4 (3/6) | <1.1 (0/4) |
| Lung | 3.4 (3/4) | 3.4 (2/4) | 4.3 (3/5) | 4.1 (1/4) |
Virus titers were based on PCR values from tissues, and the titers in some tissues were determined to generate corresponding equivalent numbers of PFU. Day 3 represents early/mild disease, and day 7 represents late/severe disease. No significant differences in virus dissemination titers were observed early in the disease; however, late in infection large differences in virus dissemination titers were observed. The limit of detection was 1.1 log10 PFU equivalent/mg tissue. The number of samples positive for virus/total number of samples assayed for each time point are noted in parentheses.
Cellular responses to infection.
We then analyzed the histology of the primary lesions and performed complete blood counts (CBCs) of rabbits infected by the two routes of infection. The histopathologies of the primary lesions in the i.d. infected rabbits and scarified lesions of virus- and PBS-infected animals were compared every 24 h from 1 to 3 days after infection (Fig. 4). Scarification with virus or without virus (PBS infected) generated a significant amount of inflammation and edema by 1 dpi, suggesting that the physical act of scarification itself damages the epithelium and underlying blood vessels, facilitating the influx of nonspecific inflammatory cells into the area. This response begins to recede in the PBS-treated animals, while animals infected with VV continue to mount an immune response that consists of a mixed infiltrate at 1 dpi and progresses to a predominately monocyte infiltrate by 5 dpi (Table 2).
FIG 4.
Histology of primary lesions from animals infected with VV or mock infected with PBS by the i.d. route and scarification. Primary lesions from i.d. infected rabbits demonstrated inflammation 24 h after lesions from rabbits infected by scarification. Lesions at 2 dpi (A) and 3 dpi (B) from animals treated with PBS i.d., 2 dpi (C) and 3 dpi (D) from animals scarified with PBS, 2 dpi (E) and 3 dpi (F) from animals infected with VV i.d., and 2 dpi (G) and 3 dpi (H) from animals infected with VV by scarification are shown. Inflammatory cell infiltration into the dermis was most prominent at 2 and 3 dpi in animals scarified with VV (G and H, respectively) and at 3 dpi in animals i.d. infected with VV (F). Arrows, areas of crust formation where the integrity of the epidermis has been damaged and neutrophilic inflammation is apparent. Magnifications, ×100.
TABLE 2.
Comparison of histology of primary lesions over time after scarification or i.d. injection of VV or PBSa
| Time (h) postinoculation | Treatment group | Epidermal inflammation | Dermal inflammation | Subcutaneous inflammation |
|---|---|---|---|---|
| 4 | PBS, scarified (n = 3) | PMN (2) | PMN (2), Mono (1) | Mixed (2), PMN (1) |
| 4 | PBS, i.d. (n = 3) | PMN (1) | Mixed (3) | None (3) |
| 24 | PBS, scarified (n = 3) | PMN (2) | Mixed (2) | Mixed (2) |
| 24 | VV, scarified (n = 4) | PMN (2) | Mixed (4) | Mixed (3), PMN (1) |
| 24 | PBS, i.d. (n = 3) | PMN (1) | Mixed (1), Mono (1) | None (3) |
| 24 | VV, i.d. (n = 4) | None (4) | Mixed (2) | None (4) |
| 48 | PBS, scarified (n = 3) | PMN (2) | Mixed (1), PMN (2) | Mixed (2) |
| 48 | VV, scarified (n = 4) | PMN (4) | Mixed (4) | Mixed (4) |
| 48 | PBS, i.d. (n = 3) | PMN (1) | Mixed (2), Mono (1) | None (3) |
| 48 | VV, i.d. (n = 4) | None (4) | Mixed (4) | Mixed (3), PMN (1) |
| 72 | PBS, scarified (n = 1) | None (1) | Mono (1) | Mono (1) |
| 72 | VV, scarified (n = 4) | PMN (4) | Mixed (4) | Mixed (4) |
| 72 | PBS, i.d. (n = 1) | None (1) | Mixed (1) | None (1) |
| 72 | VV, i.d. (n = 4) | PMN (1) | Mixed (3), Mono (1) | Mixed (2), PMN (2) |
| 120 | VV, scarified (n = 5) | PMN (5) | Mixed (5) | Mixed (1), PMN (1), Mono (3) |
| 120 | VV, i.d. (n = 5) | PMN (5) | Mixed (5) | PMN (5) |
| 168 | VV scarified (n = 5) | PMN (3) | Mixed (3) | Mixed (5) |
| 168 | VV, i.d. (n = 5) | PMN (5) | Mixed (5) | PMN (5) |
Inflammation in the skin at the site of inoculation was determined. The number of animals (n) in each experimental group is indicated. The number of animals with inflammation at the site of inoculation is indicated in parentheses. Abbreviations: i.d. = i.d. inoculation; PMN, polymorphonuclear cells (i.e., neutrophils/heterophils); Mono, mononuclear cells; mixed, polymorphonuclear and mononuclear cells.
Histological analysis of the popliteal lymph nodes, liver, spleen, and lung was also performed. Major differences between the liver and lung tissues were observed (data not shown). Differences were observed in the popliteal lymph nodes and the spleen. Both immune tissues showed evidence of a more rapid response consisting of both lymphadenomegaly and increases in the numbers of follicles in the spleen of animals infected by scarification than in the spleen of animals infected intradermally by 1 to 2 days.
While the inflammatory responses in the primary lesions of the i.d. treated animals were very similar to those in the primary lesions of the scarified animals, the increased response to infection observed in the spleen and popliteal lymph nodes of scarified animals was reflected in CBCs. When the CBCs over time of i.d. infected and scarified animals normalized against those of uninfected rabbits were compared, the differences in the percentages of neutrophils and monocytes were distinct between the two groups early in infection but were not different after 3 dpi (Table 3). The scarified animals exhibited an increase in the percentage of both neutrophils and monocytes from 1 to 2 dpi, having a 2.2-fold increase in the percentage of neutrophils and a 2.4-fold increase in the percentage of monocytes, while the i.d. infected animals demonstrated no differences between 1 and 2 dpi. The percentage of monocytes continued to increase in the scarified animals, reaching a 4.7-fold increase by 3 dpi compared to that at 1 dpi, correlating well with the shift to a focused monocyte infiltrate in the primary lesion late in infection. i.d. infected animals showed an increase similar to that of the scarified animals. However, it was delayed to 3 dpi, a full day later. The percentage of neutrophils increased by 1.6 times between 1 and 3 dpi, while the percentage of monocytes increased by 2.9 times between 1 and 3 dpi. Again, this correlates well with the observations that the i.d. infected animals had a 1-day delay in mounting an immune response and exhibiting immune cell infiltrate at the site of the primary lesion.
TABLE 3.
Comparison of early global response in animals infected via scarification and i.d. injection
The early global response was measured as the percentage of neutrophils and monocytes in animals infected via scarification and i.d. injection. Scarified animals exhibited more rapid responses, with the indicated fold increases over the average normal ranges for age-matched uninfected control animals being observed 1 day earlier in scarified animals than i.d. infected animals. While all values reported here are within the normal ranges for rabbits, they are outside of the range observed in experimental age-matched uninfected control animals. Data are for ≥3 animals.
Ring scarification.
Since we noted histologically that scarification of rabbits with only PBS induced a localized immune cell infiltrate (Fig. 3), we were interested whether this nonspecific response could afford protection against intradermal infection by the virus. We addressed this question through a ring scarification experiment which allowed us to evaluate the timing and proximity of scarification required to protect against intradermal infection. The ring scarification procedure utilizes a small circular area where the i.d. infection is introduced into the center of the circle (defining a bulls-eye), and scarification is carried out on the outer rim of the defined circle. Both scarification followed by i.d. infection and i.d. infection followed by scarification were performed, with the second procedure being performed as rapidly as possible (within approximately 2 min). In the subset of experiments shown here, all scarifications were done in the absence of virus using only PBS. Ring scarification alone was also performed using PBS to determine any unanticipated pathology.
Animals that received scarification before the i.d. injection had a survival rate of 72.7%, whereas animals that received the i.d. injection followed by scarification had a survival rate of 16.7% (Fig. 5B). Primary and secondary lesion formation, body temperatures, and weight loss were nearly identical in the two groups; however, the overall clinical scores for animals receiving scarification before i.d. injection were lower than those for animals receiving the i.d. injection before scarification (Fig. 5C), indicative of less severe disease. The difference in clinical scores was attributed primarily to the respiratory symptoms, in that animals that received scarification prior to i.d. injection exhibited fewer and less severe respiratory symptoms. It is clear that the scarification process itself prior to intradermal infection was protective.
FIG 5.
Ring scarification in combination with intradermal injection. Scarification was performed in the absence of virus. Virus was introduced during intradermal infection. (A) Photograph of the injection/scarification site 30 min after scarification/intradermal injection. The blue marker represents the site in which the scarification (red ring in center) occurred; intradermal infection occurred inside the scarification ring. (B) Survival (P = 0.0564). (C) Average clinical score over time. Error bars represent SEMs. Data for PBS-treated animals were from a historical collection of data for animals intradermally injected with PBS. There were no statistically significant differences for any measure. Six rabbits were used for i.d. inoculation and then scarification, and 12 rabbits were used for scarification and then i.d. inoculation. ⧫, PBS treatment; ▲ i.d. inoculation and then scarification; ×, scarification and then i.d. inoculation.
We next sought to determine whether there is any distal protective response or whether protection is strictly localized and whether scarification with added virus had any protective benefits, i.e., whether scarification at a site on the opposite side of the animal can protect from lethal i.d. infection with VV. Following scarification on one flank with VV or with PBS, the animals were immediately i.d. infected with virus on the contralateral flank or vice versa. Scarification with PBS was used as the control treatment. All animals that received VV via scarification on the contralateral flank, whether it was before or after the i.d. injection of VV, survived the infection (data not shown.). This was not surprising, given that it is well documented that persons can be vaccinated for smallpox several days postexposure and survive. So, by scarification on the contralateral flank, we were merely vaccinating the rabbits. A more interesting finding was that when the scarification on the contralateral flank was performed with PBS in the absence of virus, no rabbits survived and the disease progression was identical to that in rabbits with normal i.d. infection with VV. We conclude from the results of these protection experiments that it is localized scarification that generates protection against the intradermal infection and that to be protective, local scarification must precede intradermal infection.
Microarray analysis of primary lesions.
On the basis of the data demonstrating the differences in the responses of the animals from the ring scarification experiments, it was clear that virus-independent protection is initiated within the first 2 min following scarification. Microarray analysis of the primary lesions was then performed to characterize the primary immune responses responsible for this phenotype. Primary lesions were isolated at 4 h postinfection (hpi) and 1 and 2 dpi, and the RNA was isolated and analyzed using a microarray specific for rabbit genes created by the University of Florida in conjunction with Agilent. This array has previously been used to analyze the differences observed in herpesvirus infections of NZW rabbits (19).
The resulting heat map of probes identified to be significantly different between the groups (see Table S1 in the supplemental material) is shown in Fig. 6, which shows a clear difference in the i.d. infected and scarified animals by 4 hpi, as expected on the basis of the clinical data. At 4 hpi, the scarified group exhibited a response that was almost a mirror image of that of the i.d. infected group, suggesting that the responses that dictate survival of the infection are in place early in infection. The expression patterns between the groups changed at 2 dpi, with a nearly opposite expression pattern being observed between the i.d. infected and scarified groups at this time point compared to the expression pattern seen at 1 dpi. Each time point was also analyzed separately to assess whether at a given time point there were low-level transcript differences that were not identified during the large analysis, thereby identifying any transcripts exhibiting statistically significant changes.
FIG 6.
Heat map of probes identified to be class predictors over time. Results are for dermal lesion samples compared to untreated rabbit skin samples obtained at 2 dpi. The rapid response of the scarified group is demonstrated, with the majority of the probes being downregulated (blue) at 4 hpi and then rapidly upregulated (orange) at 2 dpi. The i.d. infected group did not demonstrate a downregulation of these transcripts until 2 dpi. Blue, downregulation of selected probes; black, no change from normal; orange, upregulation of selected probes; gray, missing values.
Using the same criteria for normalization and significance, analysis of the data set for 4 hpi identified 839 probes that were significant. One hundred twenty probes representing 57 genes were identified to be significant for class prediction at the 0.001 significance level (see Table S2 in the supplemental material) This class prediction was highly significant, with the probability of incorrectly assigning any array being <0.001 (data not shown). The genes were focused on skin and muscle repair, which was not surprising, given the level of tissue damage created with the act of scarification compared to that achieved with an intradermal injection. We believe that these were nonspecific responses which are independent of the presence of virus; however, they were sufficient to predict clinical outcomes via class prediction.
Analysis of the samples obtained at 1 dpi identified 927 probes to be significant. However, class prediction was not possible, as only 4 probe sets representing 3 genes were identified to be significant during the class prediction analysis. This time point exhibited a great deal of noise and no clear trends for changes in transcript levels within the treatment groups. This is thought to be due to the changes in immune responses that occur at this time point, making 1 dpi a transitional time in the local reaction. The genes identified included those for coagulation factor II, interleukin-1 (IL-1) receptor antagonist, and S100 calcium binding protein A9. All 3 genes exhibited higher expression levels in scarified animals than i.d. infected animals. These results are proposed to be due to the healing and repair response at the site of tissue damage in the scarified animals at this time point.
By 2 dpi the noise within the groups observed at 1 dpi was no longer present. There were 769 probes that were identified to be significant, and 75 probes were identified to be significant for class comparison. Upon class prediction analysis, there were 38 probes representing 29 genes that were identified to be significantly different between the classes at a 0.001 significance level (see Table S3 in the supplemental material). At 2 dpi, there was a clear separation of the two groups in the magnitude of the inflammatory response in histologic sections from the site of infection.
Likewise, the majority of the genes identified in the class prediction in the data for 2 dpi were components of the immune response, in which scarified animals had an increase in expression levels and i.d. infected animals exhibited expression levels at or near those at the baseline. Genes involved in chemokine signaling (chemokine ligands 4 and 2), complement cascade (complement component 4 and the R subunit of component 1), antigen presentation (major histocompatibility complex class II), antibody production (immunoglobulin heavy chains, variable domains, and kappa chains) and general inflammation (RIG-I, interleukin-1β, interleukin-6, interleukin-8, tumor necrosis factor, serum amyloid A2, and myxovirus resistance 2) were detected. Given that genes encoding immunoglobulin chains are upregulated, it is clear that the immune response has a strong antibody component in scarified animals at 2 dpi. Indeed, both cellular and humoral responses to VV infection are well documented following vaccination with VV (21). The genes determined to be predictors of class at 2 dpi were selected at all time points to assess the trends over time in a heat map (Fig. 6).
DISCUSSION
The vastly different outcomes for rabbits infected via either the i.d. or scarification route of infection with VV is clearly dependent on the route of infection. Animals infected via the i.d. route succumb to a lethal infection within 8 days, while animals infected via scarification survive infection. While this study used 1,000 PFU for the i.d. route and approximately 107 PFU for scarification (the needle dipped into a solution of virus at a concentration of 109 PFU/ml holds ∼10 μl), the same observations hold true for higher and lower doses of infection via both routes. i.d. inoculated animals succumbed to lethal disease even when infected with 100 PFU, and all of the scarified animals survived with either milder or no disease when lower concentrations of virus were used for inoculation. The difference in survival was not due to impairment of virus dissemination in scarified animals, for scarified and i.d. infected animals exhibited similar virus titers in tissues at 3 dpi.
The most surprising finding in these studies was that the act of scarification itself at the infection site (ring scarification), even in the absence of virus, conferred significant protection against lethal i.d. infection with either VV or RPV (data not shown). Histological examination of the lesions, global blood count changes, and rabbit-specific microarrays allowed us to examine the inoculation site, and collectively, these studies offer insight into the early response of the animals, which occurs rapidly following infection, as well as the nature of the induced immune response (19). Activation of the immune response, as reflected by CBC analysis and histology, occurred in scarified animals 24 h before activation in intradermally infected animals.
Lesions from animals scarified even in the absence of virus exhibited histology and induction of genes consistent with a very rapid nonspecific immune response at the inoculation site. The induction of these genes occurred a day earlier in lesions of scarified animals than in the inoculation site/lesions created through intradermal infection. These results imply that the skin contains a complex immunoprotection sensing system(s) where triggering of this response is localized and is dependent on both the complex architecture of the skin and the subsequent response of specific cells.
The availability of rabbit-specific microarrays contributed significantly to the ability to determine the specific differences that are induced early and that ultimately lead to an effective immune response (19). The physical difference between the two routes was primarily the nonspecific localized tissue damage caused by the physical act of scarification, which led to a nonspecific response in scarified animals, as reflected in the histology. This nonspecific response was also reflected in the microarray data by 4 h postinfection and was completely independent of the presence of virus. Increases in the levels of expression of genes associated with remodeling and repair of damaged tissue were observed at 4 hpi. It is this nonspecific, immediate response to tissue damage induced by scarification that sufficiently primes the local tissue as well as the immune system to act against the virus infection. This effect was nearly immediate (within 2 min), but it remains to be determined how long this protective response lasts.
The mechanism that we propose for this protection is illustrated in Fig. 7. We propose that the protection offered by scarification is induced by damage to the skin, which leads to a protective immune response initiated by keratinocytes. The model is based upon published data on skin function as well as our microarray data from scarified animals (22, 23).
FIG 7.
Working model of the primary local response to scarification for virus sensing and subsequent immune responses contributing to host survival. The act of scarification damages keratinocytes and endothelial cells, inducing the activation of platelets, an immediate release of intracellular stored cytokines, and induction of the inflammatory response. The immediate protective inflammatory response recruits immune cells to the site of virus delivery in scarified animals 24 to 48 h earlier than in i.d. infected animals. The local response at the cellular level creates an antiviral state, reducing the size of the primary lesion, while the global upregulation of immune-specific responses, including fever and cellular infiltrates, aids with accelerating the control of virus replication at distal sites.
We propose that following scarification, keratinocytes immediately secrete preformed, stored chemokines such as tumor necrosis factor alpha, S100 calcium binding protein A9, and IL-1β, thereby inducing an inflammatory response that is present before the virus induces an actual productive infection. The release of cytokines also induces a local response in the epidermal and dermal cells and a general antiviral state within the cells. This is an important feature, given that upon infection poxviruses are quite proficient at controlling the host immune response through the subsequent induction of a number of virus host defense genes. Keratinocytes also act as phagocytic cells, taking up virus particles, damaged cell debris, and other foreign substances, such as pathogens in the skin that may be present after the act of scarification.
There is a general increase in antigen presentation by both keratinocytes and other immune cells that have been recruited to the site of infection. The leukocytes aid in the control of the virus infection at the primary site of inoculation, promoting the healing of the site by destruction of infected cells and clearing of virus and cell debris. There is a general activation of the complement pathways to aid with the control of the virus infection. It has previously been documented that complement (observed to be upregulated in the scarified animal microarray data) is required to control virus dissemination within the animal, which supports our observation that virus dissemination was, in fact, controlled late in infection (24). It is clear that in hosts infected with poxviruses, the route of infection of the skin is a major factor contributing to protection, as it is the route of infection that dictates the response and the ultimate outcome of the disease.
The process of scarification, which promotes an important, although nonspecific, immune response against VV, could be viewed as treatment with an adjuvant. Our results suggest that it would be worthwhile to explore nearby scarification or abrading of the skin when other vaccines are injected, as significant immune enhancement and protection might result. This is a future avenue of investigation well worth exploring for vaccines in which the antigen does not evoke a strong immune response for protection alone. At least for poxviruses, it is clear that both humoral and cellular responses are induced as a result of infection via scarification. Given this example, it is possible that scarification in some cases could replace the use of chemical adjuvants. Modulating the immune response with something as simple as scarification has the potential to impact humanity and change the design of vaccines.
Supplementary Material
ACKNOWLEDGMENTS
This work was supported by National Institutes of Health (NIH) grant AI 15722 and NIH grant U54 AI057157 from the Southeastern Regional Center of Excellence for Emerging Infections and Biodefense.
The contents of this article are solely the responsibility of the authors and do not necessarily represent the official views of the NIH.
We thank David Bloom and Gregory Schultz at the University of Florida for their thoughtful discussions and guidance on the use of the rabbit-specific microarray platform.
Footnotes
Published ahead of print 23 April 2014
Supplemental material for this article may be found at http://dx.doi.org/10.1128/JVI.00185-14.
REFERENCES
- 1.Fenner F. 1989. Risks and benefits of vaccinia vaccine use in the worldwide smallpox eradication campaign. Res. Virol. 140:465–466 [DOI] [PubMed] [Google Scholar]
- 2.Gammon DB, Snoeck R, Fiten P, Krecmerová M, Holý A, De Clercq E, Opdenakker G, Evans DH, Andrei G. 2008. Mechanism of antiviral drug resistance of vaccinia virus: identification of residues in the viral DNA polymerase conferring differential resistance to antipoxvirus drugs. J. Virol. 82:12520–12534. 10.1128/JVI.01528-08 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Prichard MN, Kern ER. 2005. Orthopoxvirus targets for the development of antiviral therapies. Curr. Drug Targets Infect. Disord. 5:17–28. 10.2174/1568005053174627 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Jacobs BL, Langland JO, Kibler KV, Denzler KL, White SD, Holechek SA, Wong S, Huynh T, Baskin CR. 2009. Vaccinia virus vaccines: past, present and future. Antiviral Res. 84:1–13. 10.1016/j.antiviral.2009.06.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Phelps A, Gates AJ, Hillier M, Eastaugh L, Ulaeto DO. 2005. Comparative efficacy of replicating smallpox vaccine strains in a murine challenge model. Vaccine 23:3500–3507. 10.1016/j.vaccine.2005.02.005 [DOI] [PubMed] [Google Scholar]
- 6.Rees DGC, Gates AJ, Green M, Eastaugh L, Lukaszewski RA, Griffin KF, Krieg AM, Titball RW. 2005. CpG-DNA protects against a lethal orthopoxvirus infection in a murine model. Antiviral Res. 65:87–95. 10.1016/j.antiviral.2004.10.004 [DOI] [PubMed] [Google Scholar]
- 7.Edghill-Smith Y, Golding H, Manischewitz J, King LR, Scott D, Bray M, Nalca A, Hooper JW, Whitehouse CA, Schmitz JE, Reimann KA, Franchini G. 2005. Smallpox vaccine-induced antibodies are necessary and sufficient for protection against monkeypox virus. Nat. Med. 11:740–747. 10.1038/nm1261 [DOI] [PubMed] [Google Scholar]
- 8.Hooper JW, Thompson E, Wilhelmsen C, Zimmerman M, Ichou MA, Steffen SE, Schmaljohn CS, Schmaljohn AL, Jahrling PB. 2004. Smallpox DNA vaccine protects nonhuman primates against lethal monkeypox. J. Virol. 78:4433–4443. 10.1128/JVI.78.9.4433-4443.2004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Adams MM, Rice AD, Moyer RW. 2007. Rabbitpox virus and vaccinia virus infection of rabbits as a model for human smallpox. J. Virol. 81:11084–11095. 10.1128/JVI.00423-07 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Hutson CL, Olson VA, Carroll DS, Abel JA, Hughes CM, Braden ZH, Weiss S, Self J, Osorio JE, Hudson PN, Dillon M, Karem KL, Damon IK, Regnery RL. 2009. A prairie dog animal model of systemic orthopoxvirus disease using West African and Congo Basin strains of monkeypox virus. J. Gen. Virol. 90:323–333. 10.1099/vir.0.005108-0 [DOI] [PubMed] [Google Scholar]
- 11.Sroller V, Ludvíková V, Maresová L, Hainz P, Německová S. 2001. Effect of IFN-gamma receptor gene deletion on vaccinia virus virulence. Arch. Virol. 146:239–249. 10.1007/s007050170172 [DOI] [PubMed] [Google Scholar]
- 12.Tartaglia J, Perkus ME, Taylor J, Norton EK, Audonnet JC, Cox WI, Davis SW, van der Hoeven J, Meignier B, Riviere M. 1992. NYVAC: a highly attenuated strain of vaccinia virus. Virology 188:217–232. 10.1016/0042-6822(92)90752-B [DOI] [PubMed] [Google Scholar]
- 13.Smith SA, Kotwal GJ. 2002. Immune response to poxvirus infections in various animals. Crit. Rev. Microbiol. 28:149–185. 10.1080/1040-840291046722 [DOI] [PubMed] [Google Scholar]
- 14.Fenner F, Henderson DA, Arita I, Jezek A, Ladnyi ID. 1988. Smallpox and its eradication. World Health Organization, Geneva, Switzerland [Google Scholar]
- 15.Condit RC, Motyczka A. 1981. Isolation and preliminary characterization of temperature-sensitive mutants of vaccinia virus. Virology 113:224–241. 10.1016/0042-6822(81)90150-1 [DOI] [PubMed] [Google Scholar]
- 16.Condit RC, Motyczka A, Spizz G. 1983. Isolation, characterization, and physical mapping of temperature-sensitive mutants of vaccinia virus. Virology 128:429–443. 10.1016/0042-6822(83)90268-4 [DOI] [PubMed] [Google Scholar]
- 17.Rice AD, Adams MM, Lampert B, Foster S, Robertson A, Painter G, Moyer RW. 2011. Efficacy of CMX001 as a prophylactic and presymptomatic antiviral agent in New Zealand white rabbits infected with rabbitpox virus, a model for orthopoxvirus infections of humans. Viruses 3:63–82. 10.3390/v3020063 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Rice AD, Adams MM, Wallace G, Burrage AM, Lindsey SF, Smith AJ, Swetnam D, Manning BR, Gray SA, Lampert B, Foster S, Lanier R, Robertson A, Painter G, Moyer RW. 2011. Efficacy of CMX001 as a post exposure antiviral in New Zealand White rabbits infected with rabbitpox virus, a model for orthopoxvirus infections of humans. Viruses 3:47–62. 10.3390/v3010047 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Clement C, Popp MP, Bloom DC, Schultz G, Liu L, Neumann DM, Bhattacharjee PS, Hill JM. 2008. Microarray analysis of host gene expression for comparison between naïve and HSV-1 latent rabbit trigeminal ganglia. Mol. Vis. 14:1209–1221 [PMC free article] [PubMed] [Google Scholar]
- 20.Stern RJ, Thompson JP, Moyer RW. 1997. Attenuation of B5R mutants of rabbitpox virus in vivo is related to impaired growth and not an enhanced host inflammatory response. Virology 233:118–129. 10.1006/viro.1997.8556 [DOI] [PubMed] [Google Scholar]
- 21.Cohen JI, Hohman P, Fulton R, Turk S-P, Qin J, Thatcher K, Hornung RL. 2010. Kinetics of serum cytokines after primary or repeat vaccination with the smallpox vaccine. J. Infect. Dis. 201:1183–1191. 10.1086/651453 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Liu L, Xu Z, Fuhlbrigge RC, Peña-Cruz V, Lieberman J, Kupper TS. 2005. Vaccinia virus induces strong immunoregulatory cytokine production in healthy human epidermal keratinocytes: a novel strategy for immune evasion. J. Virol. 79:7363–7370. 10.1128/JVI.79.12.7363-7370.2005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Suter MM, Schulze K, Bergman W, Welle M, Roosje P, Müller EJ. 2009. The keratinocyte in epidermal renewal and defence. Vet. Dermatol. 20:515–532. 10.1111/j.1365-3164.2009.00819.x [DOI] [PubMed] [Google Scholar]
- 24.Cohen ME, Xiao Y, Eisenberg RJ, Cohen GH, Isaacs SN. 2011. Antibody against extracellular vaccinia virus (EV) protects mice through complement and Fc receptors. PLoS One 6:e20597. 10.1371/journal.pone.0020597 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.