Materials and Methods

Animals and sample collection—Blood samples from pet rabbits were requested from veterinarians within the United States, who submitted serum or heparinized plasma to the laboratory of the Division of Comparative Pathology at the University of Miami Miller School of Medicine. Submitted samples were accompanied by a completed questionnaire that provided information on age, sex, and clinical diagnosis. Additional samples were obtained from the Wildlife Care Center in Fort Lauderdale, Fla, as part of routine health evaluations of rabbits that were available for adoption. When received, samples were immediately processed for PE. Residual portions of samples were refrigerated at 4°C until the ELISA was performed.

Definition of clinical groups—On the basis of information from the questionnaire, presumptive diagnostic groups were formed: clinically normal rabbits, rabbits suspected of having an ECUN infection (ECUN suspect), and clinically abnormal rabbits not suspected of having an ECUN infection (non–ECUN suspect). To be classified as ECUN suspect, rabbits were required to have been suspected of having an ECUN infection, to have been treated for the infection by a participating veterinarian, and to have responded to treatment. To be classified as non–ECUN suspect, rabbits were required to be clinically ill but not believed to be infected with ECUN and not treated for ECUN infection. The non–ECUN-suspect group included rabbits with clinical signs consistent with ECUN infection but that were actually diagnosed with and treated for other infections or diseases (eg, rabbits with a head tilt or inner ear bacterial infection).

Serum or plasma PE—Fresh serum or plasma samples were analyzed with a commercial electrophoresis system^a and electrophoretic gels^b in accordance with the manufacturer's instructions. Briefly, a 1:4 dilution of each sample was overlayed on the gel and then electrophoresed for 37 minutes at 100 V. After fixation, the gel was dried and stained with blue stain reagent. After washing and drying, the gel was scanned with a densitometer to yield an electrophoretogram that quantitated each fraction as a percentage.^c Absolute values for each fraction were obtained by multiplying the percentages by the total protein concentration and are expressed as grams per deciliter. Total protein concentration was determined via refractometry.

Serum or plasma ELISA—Commercially available ELISA plates^d were used. Plates had been coated with a rabbit strain of ECUN^e that had been propagated in MRC-5 human lung fibroblast cells. Antigen was extracted via SDS treatment, heating, and centrifugation and adjusted to 3 μg/mL for plating. The ELISA plate was constructed to include 2 wells/sample: one with ECUN antigen and a second with a negative control antigen (uninfected tissue culture cells). Serum or plasma samples were serially diluted with PBS starting at a dilution of 1:32, and 50 μL of each dilution was added to each well.

After incubation at 37°C for 40 minutes, the plate was washed with a solution of saline (0.9% NaCl) and 0.05% Tween 20^f by use of a microplate washer.^g Diluted horseradish peroxidase–conjugated anti-rabbit IgG^f (50 μL) was added, and the plate was incubated for an additional 30 minutes. After plates were washed with the Tween solution, 50 μL of 2,2′-azino-bis(3-ethylbenzthi-azoline-6-sulphonic acid)^h was added. Ten minutes later, degree of absorbance at 405 nm was measured with a microplate reader.ⁱ An index was calculated by dividing the degree of absorbance of the positive antigen well by that of the negative control well. An index of ≥ 1.5 was considered a positive reaction. No differences in reactivity were detected when paired serum and plasma samples were used. Samples from confirmed ECUN-positive and ECUN-negative rabbits were concurrently evaluated in each ELISA plate to serve as control samples. Titers are reported as the last sample dilution that resulted in a positive test reaction.

Statistical analysis—A statistical software package^j was used for all analyses. Categoric data are reported as frequency and percentage, and continuous data are reported as mean ± SEM. Differences between clinical groups with respect to categoric variables were assessed by means of a C² test and logistic regression. Differences with respect to continuous variables were assessed by means of ANOVA with planned contrasts to identify differences between pairs of groups. A value of P < 0.05 was considered significant for all analyses.

Results

Animals—Veterinarians submitted plasma or serum samples from 67 clinically normal rabbits, 103 ECUN-suspect rabbits, and 33 non–ECUN-suspect rabbits. Of the rabbits classified as ECUN suspect, only 5 ECUN infections were confirmed histologically. The ECUN-suspect group included 46 rabbits with reported neurologic signs (including head tilt, ataxia, and seizures), 7 with renal signs, 15 with ocular signs, 9 with > 1 clinical sign, and 26 with various mild, nonspecific signs.

Distribution of sex within each group of rabbits was as follows: clinically normal, 64% female and 36% male; ECUN suspect, 41% female and 59% male; and non–ECUN suspect, 31% female and 69% male. Proportions of female and male rabbits in the clinically normal group were significantly (P = 0.003 for both comparisons) different from those of the ECUN-suspect and non–ECUN-suspect groups. Mean age in the 3 groups ranged between 4.3 and 4.8 years and was not significantly different among groups.

ELISA—Encephalitozoon cuniculi–suspect rabbits were more likely to be seropositive for IgG against ECUN than clinically normal and non–ECUN-suspect rabbits. Whereas 79% of the ECUN-suspect rabbits were seropositive, a significantly lower percentage of clinically normal rabbits (41%; P < 0.001) and non–ECUN-suspect rabbits (54%; P = 0.008) were seropositive for antibody against ECUN. In addition, the ECUN-suspect rabbits had significantly higher antibody titers than clinically normal non–ECUN-suspect rabbits (Table 1). The ECUN-suspect rabbits also differed from clinically normal rabbits in that ECUN-suspect rabbits had a lower A/G ratio and total protein, albumin, and A2-globulins fractions and a higher γ-globulins fraction. Similar results were evident for non– ECUN-suspect versus clinically normal rabbits. In addition, the non–ECUN-suspect rabbits had a significantly higher fraction of β-globulins than clinically normal and ECUN-suspect rabbits.

Seroprevalence of ECUN infection—Sixty-two percent of all pet rabbits evaluated were seropositive for IgG against ECUN. When result analysis was limited to rabbits obtained from a local adoption center (n = 97), the seroprevalence of ECUN infection was only 42%.

Discussion

In the present study, rabbits suspected of having an ECUN infection were more likely to be seropositive for antibody against ECUN than clinically normal rabbits. This finding was consistent with findings in some reports^10,11; however, a study¹⁸ performed in Italy revealed no significant differences in serostatus between rabbits with and without clinical signs of ECUN. We expanded on other research by determining a specific IgG titer rather than simply classifying results as positive or negative.

Quantitation of antibody against ECUN has been performed in studies of experimentally¹³ and naturally²³ infected laboratory rabbits. In those studies, high titers of antibody, often exceeding 1:1,600, could still be detected several weeks after initial infection. In another study,¹³ experimental IV inoculation of 6 rabbits with ECUN failed to yield an association between antibody response and recovery of ECUN organisms from brain tissue. In contrast, the present study revealed that rabbits suspected of having an ECUN infection had an antibody titer that was 1.7 times as high as that of clinically normal rabbits or clinically abnormal rabbits that were not suspected of having an ECUN infection. A mean titer of 1:1,324 was detected in ECUN-suspect rabbits, and many rabbits had titers in the 1:3,200 to 1:6,400 range. Such high titers of IgG against ECUN were consistent with the findings of another research group who reported the development of sustained high titers of antibody against ECUN in rabbits, lasting 7 years after inactivated spores were injected IV.¹⁴ The longest period of seropositivity after experimental ECUN infection is reportedly 400 days.⁹ In the present study, several rabbits from an adoption center were examined over 9 months. During this period, most rabbits had stable titers of IgG against ECUN, although some rabbits with initially high titers had a 2-fold decrease in antibody (data not shown). Additional studies are needed to identify the mean duration of seropositivity in rabbits with a naturally acquired ECUN infection.

In the present study, the sample of ECUN-suspect rabbits consisted predominantly of pet rabbits with neurologic signs. Because of the low number of rabbits with ocular and renal signs of infection, we were precluded from performing statistical analyses to compare clinical signs with test results. Additional studies with a greater proportion of rabbits in each classification of clinical signs are needed to explore that potential association.

We did not detect any difference between clinically normal and ECUN-suspect rabbits with respect to age. In a study¹⁸ of serostatus for ECUN infection without regard to clinical status, pet rabbits < 4 months old were reportedly more likely to be seronegative than older rabbits. In contrast, another study¹⁷ revealed no such difference. Naturally infected laboratory rabbits begin to develop maternally acquired antibody at 4 weeks of age.²³ At 8 to 10 weeks of age, titers of antibody against ECUN exceed 1:800. These findings provide evidence of the difficulty encountered when attempting to diagnose ECUN infection in young rabbits by means of serologic test. Because most rabbits in the present study were > 1 year old, an association between age and ECUN infection could not be adequately explored.

The ECUN-suspect group contained a significantly greater proportion of male rabbits than the clinically normal group. Another study⁹ of 119 seropositive rabbits from the UK revealed a similar overrepresentation of ECUN infection in male versus female rabbits; however, the difference between sexes was not significant. A smaller study¹⁷ of 52 rabbits revealed no differences in susceptibility to ECUN infection between sexes. That said, results of the present study suggest that sex of rabbit should be considered in the diagnosis of ECUN infection.

In comparison with the performance of other serologic assays, the ELISA we used is reportedly not as sensitive in the quantitation of antibody.¹⁶ The reported lack of sensitivity, however, may have been attributable to a reliance on optical density readings from only 1 serum dilution. On the other hand, in the present study, all samples were tested by serial dilution to determine the endpoint of their reactivity.

Protein electrophoresis is used as an accessory diagnostic and prognostic tool in veterinary medicine.^19–21 In ECUN-infected dogs, increased serum concentrations of total IgG as well as ECUN-specific IgG have been detected.²² In the present study, a significant increase in serum or plasma γ-globulins concentration was detected in ECUN-suspect rabbits. The present study also provided evidence of a concomitant decrease in serum or plasma A/G ratio and concentrations of albumin and A2-globulins. Albumin is generally referred to as a negative acute-phase protein (protein that decreases during the acute-phase response). Therefore, we expected the decrease in serum or plasma albumin concentration that was detected in ECUN-infected rabbits. The origin of the decrease in the serum or plasma concentration of α-globulins is unknown because most acute-phase proteins in the α-globulins fraction are believed to be positive acute-phase proteins (proteins that increase in concentration with the acute-phase reaction) in mammals. It is interesting that the clinically abnormal rabbits that were not suspected of having an ECUN infection had the same changes in serum or plasma A/G ratio and concentrations of γ-globulins, albumin, and A2-globulins.

The clinically abnormal rabbits that were not suspected of having an ECUN infection also had a higher serum or plasma β-globulins concentration than clinically normal or ECUN-suspect rabbits. Many non– ECUN-suspect rabbits reportedly had a bacterial infection that mimicked signs of an ECUN infection (eg, head tilt). Consequently, it appears this type of non-ECUN bacterial infection results in significant increases in serum or plasma concentrations of both β-globulins and γ-globulins rather than just an increase in γ-globulin concentration, as was detected in ECUN-suspect rabbits. Because most clinical signs of ECUN infection are a result of chronic inflammatory processes, perhaps the detection of increased serum or plasma β-globulins concentration in the abnormal non–ECUN-suspect rabbits represented the result of an acute infection.

Overall, the results of the present study indicated that abnormalities in serum or plasma concentrations of proteins cannot be used alone specifically for the diagnosis of ECUN infection but can be used as accessory tools in the diagnostic process. Because plasma samples were used in the present study, changes in fibrinogen concentration may have been reflected in the increases reflected in the plasma β-globulin fraction.¹⁹ In most mammals, this fraction represents immunoglobulins and possibly also represents acute-phase proteins such as C-reactive protein, particularly in rabbits.¹⁹ Increases in concentrations of acute-phase proteins may be associated with clinical signs and tissue damage associated with ECUN infection.

The prevalence of ECUN infection in rabbits appears high in the United States and elsewhere,^9,10,17,18 and ECUN infection may account for various clinical signs detected during veterinary examination of rabbits. The potential for ECUN to be zoonotically transmitted, particularly to immunocompromised people, cannot be dismissed, and improved screening tests are needed to aid in detection of ECUN infection in rabbits.^5–8