1

The Rabies Indirect Fluorescent Antibody Test: Presence of Cross–Reactions 1

with Other Viral Encephalitides 2

3

Robert J. Rudda #

, Kim A. Applera and Susan J. Wong

b 4

5

6

7 Wadsworth Center, New York State Department of Health, Albany, NY 8

a Rabies Laboratory, Wadsworth Center, New York State Department of Health, 5668 State Farm Road, 9

Slingerlands, NY 12159 (Griffin Laboratory) 10

b Diagnostic Immunology Laboratory, Wadsworth Center, New York State Department of Health, 120 New 11

Scotland Ave, Albany, NY 12208 (David Axelrod Institute-Public Health) 12

# rjr06@health.state.ny.us, corresponding author 13

kak11@health.state.ny.us 14 sjw03@health.state.ny.us 15 16 17

Keywords: viral encephalitis, diagnosis, rabies, indirect fluorescent antibody 18

Abstract 19

The antemortem diagnosis of rabies in humans employs techniques that require 20

accuracy, speed and sensitivity. A combination of histochemical analysis, in vitro virus isolation, 21

immunologic methods and molecular amplification procedures are utilized in an effort to 22

diagnose the disease. Modern medicine now offers potentially life-saving treatment for a disease 23

that was considered invariably fatal once clinical signs develop. However, medical intervention 24

efforts require a rapid and accurate diagnosis as early in the course of clinical signs as possible. 25

The indirect fluorescent antibody (IFA) procedure on cerebral spinal fluid and serum provides a 26

rapid result but the specificity of the assay has not been well studied. Because a false positive 27

IFA could significantly affect patient treatment and outcome, it is critical to understand the 28

specificity of this assay. In this study, IFA was performed on 135 cerebral spinal fluid and serum 29

specimens taken from patients with viral encephalitis or a presumed viral infection from an agent 30

JCM Accepts, published online ahead of print on 2 October 2013J. Clin. Microbiol. doi:10.1128/JCM.01818-13Copyright © 2013, American Society for Microbiology. All Rights Reserved.

on May 31, 2018 by guest

http://jcm.asm

.org/D

ownloaded from

2

other than rabies virus. Results indicate that false-positive results can occur when interpreting 31

the rabies IFA procedure. A staining pattern morphologically similar to anti-rabies staining was 32

observed in 7 of the135 spinal fluids examined. In addition, a majority of the spinal fluids tested, 33

from patients with encephalitis, presented immunoglobulin that bound to antigens present in cell 34

culture substrate. Of marked concern was the frequent presence of cross-reactive antibodies in 35

encephalitis cases associated with West Nile and Powassan flaviviruses. Because the IFA test 36

for rabies on human specimens may result in false-positive results, it should not be used as the 37

sole basis for initiating anti-rabies treatment. 38

39

40

41

42

Introduction 43

Rapid and accurate ante mortem rabies diagnosis in humans has been an imperative for 44

palliative patient care and for treatment of individuals potentially exposed to the patient. Since 45

the Milwaukee protocol (1) was introduced as a potential life-saving treatment for human rabies, 46

the sooner the protocol is initiated the greater the chance of success. This paradigm demands 47

speed and accuracy from the rabies diagnostician. The test most likely to provide a quick rabies 48

diagnosis is the direct fluorescent antibody test (DFA) (2) performed on a nuchal skin biopsy 49

from the patient. However, since this test may be negative in earlier stages of the disease, other 50

procedures are relied upon and are carried out concurrently with the DFA. The indirect 51

fluorescent antibody test (IFA) used on cerebral spinal fluid (CSF) and serum from the rabies 52

suspect patient can yield results within a few hours. To perform an IFA, serial dilutions of serum 53

or CSF are placed onto fixed rabies-infected cultured cells. If the serum or CSF contains 54

antibodies to rabies, these antibodies will attach to rabies antigens present in the infected cell 55

on May 31, 2018 by guest

http://jcm.asm

.org/D

ownloaded from

3

substrate. A fluorescein isothiocyanate (FITC) labeled secondary antibody specific for human 56

immunoglobulins is applied and the slides are then examined by fluorescence microscopy. An 57

experienced microscopist will recognize fluorescent staining patterns indicating the presence of 58

an immune response to rabies virus. 59

The IFA is a quick and sensitive procedure. However, the specificity of the assay has not 60

been studied in detail. This study analyzed the specificity of the rabies IFA test by examination 61

of specimens from rabies-negative patients who present with an encephalitis of known or 62

unknown origin. Results indicate that the specificity of the rabies IFA is not 100%, and thus 63

should not be the sole basis for initiating rabies therapy. 64

65

66

Materials and Methods 67

Cell Culture. BHK-21 cells (C-13) (ATTC CCL10) (American Type Culture Collection, 68

Rockville, MD.) were used at passages 70 to 95. Mouse neuroblastoma cells (NA), (10) were 69

used at passages 700 to 750. Both cell lines were cultured and maintained as previously reported 70

(11). 71

Virus inoculum. The ERA strain of rabies virus (12) was utilized as the rabies antigen source in 72

the IFA procedure. The virus inoculum used to infect cells was obtained from a commercially 73

available veterinary vaccine vial (13). Prior to use in the preparation of the IFA antigen slides 74

the stock virus was passaged twice in BHK-21 cells using the media previously reported (11). 75

At second passage cell confluency the flasks were placed at -80º C overnight. Cells were thawed 76

to frozen slurry, agitated and refrozen at -80º. Upon thawing lysed cell debris was removed by 77

centrifugation at 1000G and aliquots were prepared from the supernatant for storage at -80º C. 78

on May 31, 2018 by guest

http://jcm.asm

.org/D

ownloaded from

4

79

Antigen slide preparation. Stored virus inoculum had previously been titrated to identify end-80

point values and infectivity profiles in both neuroblastoma and BHK-21 cell cultures. Virus 81

inoculum was added to trypsinized cells at a MOI and cell-count suitable to produce 40 to 50% 82

cell infection with three days growth at 34º C with 5% CO

2 atmosphere in a moist chamber 83

incubator. Cells were grown on multi-well teflon coated slides (Cel-Line/Thermo Fisher 84

Scientific, Cat no. 30-225H). After 3 days of cell growth medium was removed, cells washed 85

once (2 min.) in phosphate buffered saline, 0.01M, pH 7.6 (PBS) and air dried before storage at -86

80ºC. Upon use, antigen slides were thawed, air dried, fixed in -20

º C acetone overnight and air 87

dried prior to the addition of sera/CSF. 88

89

Anti-human IgG and IgM antibodies. Goat anti-human IgG-FITC (Catalog # IF0001) was 90

obtained from Focus Diagnostics (Cypress, CA 90630). This secondary antibody is used directly 91

from the vial, with no further dilutions or additions. Goat anti-human IgM(µ)- FITC was 92

obtained from Kirkegaard & Perry Laboratories, (Gaithersburg, MD 20879), reconstituted as 93

directed by the manufacturer, and then diluted 1:40 in .01M PBS, Ph 7.6 containing Evans Blue 94

counterstain at 0.00125%. 95

96

Clinical Material. A total of 135 CSF samples from viral encephalitis patients were tested by 97

the rabies IFA procedure. The sample set included ten cases of Epstein Barr virus(EBV), one 98

eastern equine encephalitis(EEE) , one human herpesvirus 6 (HHV6) , one dual 99

HHV6+enterovirus , four enterovirus, six herpes simplex virus(HSV2), and one dual EBV/ 100

varicella zoster virus(VZV), all confirmed by real-time PCR on CSF (16). Thirty CSF samples 101

on May 31, 2018 by guest

http://jcm.asm

.org/D

ownloaded from

5

were from patients serologically diagnosed with West Nile infections and five CSF samples were 102

serologically diagnosed as Powassan encephalitis cases. The IgG and IgM assays were 103

performed on rabies infected murine neuroblastoma cells (RIC) and non-infected neuroblastoma 104

cells (NRIC). On a smaller subset of specimens rabies infected and non-infected BHK-21 cells 105

were employed to compare results with the neuroblastoma cell assay. The IgG IFA was 106

performed on 135 spinal fluids and 17 sera. IgM IFA assay was performed on 115 of the spinal 107

fluids. CSF samples were tested undiluted. Sera were available from two types of encephalitis 108

cases: West Nile and Powassan which were tested as positive in cross-species plaque reduction 109

neutralization tests(14). These sera were tested in the rabies IFA at a screening dilution of 1:20. 110

For the IgM assay sera samples were IgG depleted by an initial 1:8 dilution in goat-anti human 111

IgG (GullSorb, Meridian Bioscience, Inc., Cincinnati, OH 45244). Sera for IgM assay were 112

tested at a 1:20 dilution. Rabies positive control sera was from a vaccine recipient with a titer of 113

2.0 IU and serum and CSF obtained from a human rabies case. Negative control sera were from 114

healthy individuals with no rabies virus neutralization antibodies detected. The positive control 115

serum was diluted in serial 2-fold dilutions, in PBS. A 1/32 or 1/64 serum dilution was used as a 116

control for each IFA assay. 117

118

Indirect Fluorescent Antibody Assay. Fifteen µl of CSF or diluted sera was applied per well 119

of multiwell antigen slides. All sera were tested on RIC and NRIC for comparison. Slides are 120

incubated at 37º

C for 30 min in a moist chamber. The non-bound antibody was eluted by a 121

gentle wash with PBS (.01M, Ph 7.6) from a wash bottle and the slides were then soaked for 15 122

min at room temperature in a PBS filled coplin jar. Fifteen µl of anti-human IgG was applied to 123

each well and the slide incubated at 37º C for 30 min. After the second incubation conjugate was 124

on May 31, 2018 by guest

http://jcm.asm

.org/D

ownloaded from

6

gently washed with PBS from a wash bottle before the slides were soaked in PBS for 15 min at 125

room temperature. Slides were air dried and coverslips mounted with a mountant consisting of 126

0.05M Tris buffered 0.15 saline, pH 9.0 with 20% glycerol. IgM testing was performed as the 127

IgG assay with the substitution of goat anti-human IgM-FITC conjugate. All slides were 128

evaluated by RJR, and pertinent images were simultaneously viewed on a computer monitor by 129

SJW for discussion and grading of the reaction intensity. 130

131

Microscopy and Imaging. Photomicroscopy was performed using a Zeiss AxioImager A1 132

microscope equipped for fluorescence microscopy. A Zeiss AxioCam MRc camera captured 133

images using Zeiss AxioVision 3.1 software. Images were optimized for brightness and contrast, 134

using Adobe Photoshop Elements 2.0 software. 135

136

Rabies virus neutralization assay. CSF and serum samples that presented a structural pattern 137

of staining similar to staining pattern observed with known rabies positive cases were further 138

examined using an in-vitro rabies virus neutralization assay. (15) 139

140

Results 141

Staining patterns. The attachment of antibodies, as reflected by the attachment of anti-human 142

IgG FITC labeled conjugate, produced either structurally specific patterns or a generalized 143

background staining. A staining pattern appearing similar to specific anti-rabies staining was 144

observed in 7 of the 135 spinal fluids examined for IgG specific for rabies antigen. This 145

staining pattern was present on RIC and absent on NRIC. The IFA procedure examining IgM 146

antibodies identified 2 out of 115 spinal fluids with a rabies-like immunoreactive pattern. 147

on May 31, 2018 by guest

http://jcm.asm

.org/D

ownloaded from

7

Specific staining on RIC is identified as either intracytoplasmic inclusions consisting of rabies 148

virus ribonucleoprotein or membrane associated rabies glycoprotein (Fig. 1A,B). All of the CSF 149

or serum samples that presented IFA staining similar to rabies specific staining were negative for 150

rabies virus neutralizing antibodies. Comparison of reactivity patterns between RIC and NRIC 151

identified patterns that ranged from antibody attachment seen only in rabies infected cells ( Fig. 152

1 & S1-S5) to strong reactivity in both RIC and NRIC (Fig.S6 - S8 ). In certain clinical samples 153

the morphology of the staining in the RIC, although strong, was atypical of a specific RIC 154

staining morphology (Fig. S6, S9). Samples were also identified where there was a reduced 155

reaction pattern in the NRIC when there was a very strong reaction in the RIC (Fig. S10). Of the 156

135 samples examined for IgG reactions, 70 (51.5%) reacted to RIC and 58 (42.6%) reacted to 157

NRIC. Of the 115 samples examined for IgM reactions, 30 (26%) reacted to RIC and 28 (24%) 158

reacted to NRIC. Evidence of reactivity to RIC and NRIC as compared to the etiologic agent 159

responsible for the encephalitis is presented in Table 1. Reaction patterns with Powassan 160

positive CSF in RIC and NRIC (Fig.S6) in BHK-21 cells were particularly striking, revealing 161

antibodies with a strong avidity to these cells. Neither anti-human IgG-FITC or anti-human 162

IgM-FITC produced non-specific binding to antigen slides containing RIC or NRIC. Non-163

specific attachment of immunoglobulins from encephalitic patients to RIC and NRIC was 164

frequently observed(Table 1), often presenting a staining pattern that was targeting cytoskeleton 165

or specific organelles and was discernable from a specific rabies reaction. 166

167

168

Discussion 169

on May 31, 2018 by guest

http://jcm.asm

.org/D

ownloaded from

8

We examined CSF and sera from encephalitic human patients and identified a subset that 170

presented a positive reaction in the indirect fluorescent antibody test designed for the 171

demonstration of anti-rabies antibody. When these positive samples were tested on the gold-172

standard rabies virus neutralization test the samples were negative for anti-rabies antibody. In 173

most cases an alternate etiologic agent was identified by either PCR testing of the CSF and/or the 174

identification of serum antibodies to other pathogens. The potential for false-positive results in a 175

test designed to diagnose rabies in a human is disconcerting, as rabies is noted as an invariably 176

fatal disease. Additionally, if rabies is diagnosed in a patient it likely will initiate the events 177

associated with the Milwaukee protocol (1). 178

The presence of cross reactive antibodies induced by viral infections has been well 179

documented (3-6). Srinivasappa et al (3) demonstrated molecular mimicry when monoclonal 180

antibodies, developed against numerous viral pathogens, reacted with normal tissues from mice. 181

Antibodies directed against measles virus cross react with cellular stress proteins of mammalian 182

cells infected with heterologous viruses (4). Rabies infected cell culture could produce similar 183

stress proteins that would be recognized by antibodies directed against a heterologous 184

encephalitic agent. Solid phase assays such as the IFA measure any antibody that binds to the 185

antigen source. The antigen sources in the rabies IFA are rabies infected and non-infected cell 186

cultures. The microscopist evaluating the fluorescence reaction pattern is tasked with discerning 187

the proper staining pattern associated with a positive reaction due to attachment of antibody to 188

rabies antigen in the infected cell culture. Specific staining patterns of RIC may show both 189

intracytoplasmic inclusions containing rabies ribonucleoprotein and rabies virus surface antigen 190

containing glycoprotein (7). The morphology of the staining pattern for these two antigens may 191

be differentiated by examining RIC stained with monoclonal antibodies directed against these 192

on May 31, 2018 by guest

http://jcm.asm

.org/D

ownloaded from

9

two antigens (8). The recognition of staining patterns that are specific only to rabies infections 193

should become problematic only when there are morphologically similar staining patterns 194

induced by sources other than a rabies infection. We present five instances (Fig.S1-S5) in which 195

there was a positive reaction in RIC and the NRIC presented little to no reaction. In three of 196

these cases (Fig. S1, S4 & S5) the staining pattern in the RIC could, in our opinion, be 197

interpreted as similar to staining due to a rabies positive antibody source. 198

The 2011 case definition for the diagnosis of human rabies includes the identification of 199

Lyssavirus specific antibody by indirect fluorescent antibody test or complete rabies virus 200

neutralization at 1:5 dilution in the CSF (9). This case definition strongly recommends that 201

laboratory confirmation is accompanied by a positive result with the other diagnostic techniques 202

presently in use for antemortem human rabies diagnosis. This definition nevertheless allows for 203

the diagnosis of rabies in a human based solely on the results of the IFA procedure on CSF and 204

sera samples from encephalitic patients. Because the IFA test for rabies on human specimens 205

may result in false-positive results, it should not be used as the sole basis for diagnosis of rabies 206

in humans and implementation of experimental approaches to anti-rabies treatment. Further 207

work is warranted to determine the prevalence of cross reactivity when employing solid phase 208

assays such as the IFA for the diagnosis of rabies and to validate by inter-laboratory proficiency 209

testing any such test. 210

211

Acknowledgments 212

The authors would like to acknowledge the excellent technical assistance provided by Craig 213

Pouilott, Jodie Jarvis, Kathleen Brown, Patrick Fitzgerald, Anne Clobridge and Michelle Dupuis. 214

We are indebted to Dr. Marvin Fritzler for valuable discussions concerning staining phenotypes 215

on May 31, 2018 by guest

http://jcm.asm

.org/D

ownloaded from

10

in the IFA test. We are grateful for the editing provided by Drs. Harry Taber, Ron Limberger 216

and Kirsten St. George. The Photography and Illustrations core at the Wadsworth Center was 217

instrumental in preparing the Table and Figures used in this work. We are indebted to the Viral 218

Encephalitis Laboratory and Diagnostic Immunology Laboratories at the Wadsworth Center for 219

the sharing of samples that made this study possible. 220

221

Author Contributions 222

RJR and SJW conceived and designed the experiments. RJR, SJW and KAA performed the 223

experiments. RJR and SJW wrote the paper. 224

225

226

References 227

228

1. Willoughby RE, Tieves KS, Hoffman GM, Ghanayem N, Amlie-Lefond CM, Schwabe 229

MJ, Chusid MJ, Rupprecht, CE. 2005. Survival after treatment of rabies with induction of 230

coma. New Engl. J. Med. 352: 2508–14. doi:10.1056/NEJMoa050382. 231

232

2. www.cdc.gov/rabies/pdf/RabiesDFASPv2.pdf . Accessed 27 June 2013. 233

234

3. Srinivasappa J, Saegusa J, Prabhakar BS, Gentry MK, Buchmeier MJ, Wiktor TJ, 235

Koprowski H, Oldstone MBA, Notkins AL. 1986. Molecular mimicry: Frequency of reactivity 236

of monoclonal antiviral antibodies with normal tissues. J. Virol. 57:397-401. 237

on May 31, 2018 by guest

http://jcm.asm

.org/D

ownloaded from

11

238

4. Sheshberadaran H, Norrby E. 1984. Three monoclonal antibodies against measles 239

virus F protein cross-react with cellular stress proteins. J. Virol. 52: 995-999. 240

241

5. Fujinami RS, Oldstone MA, Wroblewska Z, Frankel ME and Koprowski H. 1983. 242

Molecular mimicry in virus infections: Cross-reaction of measles virus phosphoprotein or of 243

herpes simplex virus protein with human intermediate filaments. Proc. Natl. Acad. Sci. USA 80: 244

2346-2350. 245

246

6. Crawford L, Leppard K, Lane D, Harlow E. 1982. Cellular proteins reactive with 247

monoclonal antibodies directed against simian virus 40 T- antigen. J. Virol. 42: 612-620. 248

249

7. Rupprecht C E, Hanlon CA, Hemachudha T. 2002. Rabies: re-examined. Lancet Infect. 250

Dis. 2: 327–343. 251

252

8. Zanlucaa C, Passos Airesa LR, Pazzini Muellera P, Santosa VV, Luiza Carrieri M, 253

Roberto Pintoa A, Roberto Zanettia C. 2011. Novel monoclonal antibodies that bind to wild 254

and fixed rabies virus strains. J. Virol. Methods 175: 66-73. 255

256

9.http://wwwn.cdc.gov/NNDSS/script/casedef.aspx?CondYrID=818&DatePub=1/1/2011%2012:257

00:00%20AM , Accessed 27 June 2013. 258

259

on May 31, 2018 by guest

http://jcm.asm

.org/D

ownloaded from

12

10. McMorris F A, Ruddle FH. 1974. Expression of neuronal phenotypes in neuroblastoma 260

cell hybrids. Dev. Biol. 39: 226-246. 261

262

11. Rudd RJ, Trimarchi CV. 1987. Comparison of sensitivity of BHK-21 and murine 263

neuroblastoma cells in the isolation of a street strain rabies virus. J. Clin. Microbiol. 25:1456-264

1458. 265

266

12. Abelseth M K. 1964. Propagation of rabies virus in pig kidney cell culture. Can. Vet. J. 5: 267

84-87. 268

269

13. Lawson KF, Crawley JF. 1972. The ERA Strain of Rabies. Can. J. Comp. Med. 36: 339–270

344. 271

272

14. Lindsay H, Calisher CH, Mathews JH. 1976. Serum dilution neutralization test for 273

California group virus identification and serology. J. Clin. Microbiol. 4: 503-510. 274

275

15. Trimarchi CV, Rudd RJ, Safford M Jr. 1996. An in vitro virus neutralization test for 276

rabies antibody. In: Laboratory Techniques in Rabies. Fourth Edition, WHO. Geneva. 277

278

279

16. Dupuis M, Hull R, Wang H, Nattanmai S, Glasheen B, Fusco H, Dzigua L, Markey K, 280

Tavakoli NP. 2011. Molecular detection of viral causes of encephalitis and meningitis in New 281

York State. J Med Virol. 83: 2172-81. 282

on May 31, 2018 by guest

http://jcm.asm

.org/D

ownloaded from

13

283

284

285

TABLE 1 Evidence of antibody attachment in CSF samples examined 286

with the Indirect Fluorescent Antibody Test. 287

Causative Agent of

Encephalitisa

Evidence of antibody reactivity on rabies infected

and non-infected cell cultureb

RIC NRIC

West Nile 20/30 (66.6%) 18/30 (60%)

Powassan 4/5 (80%) 2/5 (40%)

EBV 8/10 (80%) 6/10 (60%)

HSV2 3/6 (50%) 3/6 (50%)

HHV6 & Entero 1/1 (100%) 0/1 (0%)

Entero 1/3 (33%) 0/3 (0%)

HHV6 1/1 (100%) 0/1 (0%)

EBV & VZV 1/1 (100%) 1/1 (100%)

EEE 0/1 (0%) 0/1 (0%)

Unknown 45/78 (57.6%) 35/78 (44.8%)

a West Nile and Powassan cases diagnosed by serology, all others 288

diagnosed by real-time PCR and real-time RT-PCR on CSF (16) 289

b Rabies infected cells (RIC) and non rabies infected cells (NRIC) were 290

examined by the rabies indirect fluorescent antibody (IFA) procedure. 291

292

293 294

on May 31, 2018 by guest

http://jcm.asm

.org/D

ownloaded from

14

295

296

297

298 299

on May 31, 2018 by guest

http://jcm.asm

.org/D

ownloaded from

on May 31, 2018 by guest

http://jcm.asm

.org/D

ownloaded from