J Vet Intern Med 2003;17:538 544 Polymerase Chain Reaction (PCR) Amplification of Parvoviral DNA from the Brains of Dogs and Cats with Cerebellar Hypoplasia Scott J. Schatzberg, Nicholas J. Haley, Stephen C. Barr, Colin Parrish, Samantha Steingold, Brian A. Summers, Alexander delahunta, Joe N. Kornegay, and Nicholas J.H. Sharp Cerebellar hypoplasia in cats is caused most commonly by an in utero or perinatal infection with feline panleukopenia virus (parvovirus). Cerebellar hypoplasia has been reported infrequently in dogs, but no viral etiology has been identified to date. DNA was extracted from archival, paraffin-embedded, cerebellar tissue from 8 cats and from 2 canine littermates with cerebellar hypoplasia, 2 canine littermates with cerebellar cortical abiotrophy, 6 dogs with congenital cerebellar vermal defects, 1 dog with congenital hydranencephaly, and 15 dogs and cats with various encephalitides. The DNA extracted from each cerebellum was subject to polymerase chain reaction (PCR) amplification by 3 primer pairs specific for parvovirus DNA. Sequence analysis of PCR products from each of the 8 cats and 2 dogs with cerebellar hypoplasia confirmed their identity with parvoviral DNA. The 6 dogs with cerebellar vermal defects, 2 dogs with cortical abiotrophy, 1 dog with congenital hydranencephaly, and all control samples were PCR negative for parvovirus. Parvoviral structural proteins were not identified by immunohistochemistry in either dog with cerebellar hypoplasia. This study shows that parvoviral DNA can be amplified from feline and canine archival brain tissue and that cerebellar hypoplasia in dogs might be associated with in utero parvovirus infection. Key words: Abiotrophy; Canine; Cerebellum; Dandy-Walker Syndrome; Feline; Parvovirus. Congenital cerebellar abnormalities occur sporadically in domestic animals. The 2 most common disorders are hypoplasia and atrophy after in utero (or perinatal) viral infection and primary developmental defects. Viral infections occur most commonly in cats and cattle but have been described in sheep, pigs, goats, and chickens. 1 5 Cerebellar hypoplasia has not been described in the dog secondary to in utero viral infection. Congenital cerebellar malformations are reported occasionally in dogs, and typical lesions include partial or complete absence of the cerebellar vermis. 6 9 Such malformations often are described as examples of Dandy-Walker syndrome, a congenital anomaly of humans in which the primary defect is a partial or complete absence of the cerebellar vermis combined with a cystic dilatation of the roof of the 4th ventricle. The most common cause of cerebellar disease in dogs is cerebellar cortical abiotrophy. 5 With few exceptions, puppies are neurologically normal at birth but develop progressive cerebellar deficits within weeks to months. This finding is in marked contrast to what is observed in animals with viral causes of cerebellar ataxia or vermal defects in which clinical signs remain static or even improve slowly as the animal compensates for its deficit. Abiotrophies are presumed to be genetic in origin and lead to premature death of neurons in the cerebellar cortex. 5,10 From the Cornell University Hospital for Animals, Cornell University, Ithaca, NY (Schatzberg, Haley, Barr, Parrish, Summers, de- Lahunta); The Brody School of Medicine at East Carolina University, Greenville, NC (Steingold); the College of Veterinary Medicine, University of Missouri at Columbia, MO (Kornegay); and the Animal Critical Care Group of Vancouver, Burnaby, BC Canada (Sharp). The abstract for this paper was previously presented at the 2002 American College of Veterinary Internal Medicine meeting in Dallas, Texas. (May 29 June 1). Reprint requests: Dr Scott J. Schatzberg, DVM, PhD, Department of Clinical Sciences, Cornell University, CUHA, Box 35, Ithaca, NY 14853-6401; e-mail: sjs53@cornell.edu. Submitted July 30, 2002; Revised November 27, 2002; Accepted January 15, 2003. Copyright 2003 by the American College of Veterinary Internal Medicine 0891-6640/03/1704-0009/$3.00/0 Cerebellar hypoplasia, in contrast to cerebellar abiotrophy or vermal malformation, is reported uncommonly in the dog. In 1995, Summers et al briefly described a young adult Beagle dog with an unknown history that had signs of a static cerebellar disorder. 5 On postmortem examination, the dog had a moderately hypoplastic cerebellum with extensive microscopic lesions similar to those seen in cats after in utero viral infections. No etiology was identified. In 1992, Kornegay identified 2 Blue Tick Coonhound littermates that manifested cerebellar signs from the time of birth. Postmortem examination disclosed diffusely hypoplastic cerebella with no other gross malformations (unpublished observations). Tissues from these 2 dogs were available for our study, in which we hypothesize that canine cerebellar hypoplasia is caused by in utero infection with canine parvovirus. We have amplified parvoviral DNA recently from paraffinized feline brain tissue. 11 Parvoviral DNA also has been amplified previously by polymerase chain reaction (PCR) on fecal material and on a variety of fresh and fixed tissues, including mesenteric lymph node, thymus, and ileum. 11 14 In the present study, formalin-fixed, paraffin-embedded brain tissue from 8 cats and 2 canine littermates with cerebellar hypoplasia, 6 dogs with vermal malformations, 2 canine littermates with cerebellar cortical abiotrophy, and 1 dog with congenital hydranencephaly were studied for the presence of parvovirus by PCR. Materials and Methods Case Selection Ten dogs and 8 cats in this study presented to the neurology services at the College of Veterinary Medicine, North Carolina State University, or to Cornell University Hospital for Animals (CUHA) with clinical signs of cerebellar disease. One dog with congenital hydranencephaly presented to CUHA and has been described previously (case 11). 12 The cats (cases 12 19) all presented with a presumptive clinical diagnosis of cerebellar hypoplasia, which was confirmed on postmortem examination. Five of these cats were unrelated and 3 were from the same litter. Two of the 8 dogs (cases 1 and 2) have been described previously with cerebellar vermal defects of unknown etiology. 8 The 8 additional study animals included 4 vermal defects (cases 3 6), 2
Cerebellar Parvovirus 539 canine littermates with cerebellar cortical abiotrophy (cases 7 and 8), and 2, 7-month-old Blue Tick Coonhound littermates (cases 9 and 10) born with cerebellar hypoplasia. Control animals comprised 15 animals in which other CNS disorders were confirmed by histopathologic evaluation. These included 10 cats (5 with CNS toxoplasmosis, 3 with FIP encephalitis, and 2 with intracranial neoplasia) and 5 dogs (2 with encephalitis of unknown etiology and 1 each with CNS toxoplasmosis, pseudorabies encephalitis, and intracranial neoplasia). Histopathology Archival tissues that had been fixed in 10% buffered formalin and embedded in paraffin were sectioned serially at 5 6 m, deparaffinized, and stained with hematoxylin and eosin. Immunohistochemistry Immunohistochemical studies for parvovirus were performed on the formalin-fixed, paraffin-embedded specimens from 2 dogs (cases 9 and 10) and 1 cat (case 18) with ProbeOn slides. a Briefly, tissues were deparaffinized and rehydrated in 70% ethanol. Tissues were blocked with 0.5% hydrogen peroxide in methanol for 10 minutes at room temperature and reblotted and washed. For antigen retrieval, the tissues were incubated with 1 Pronase for 20 minutes at 37 C and again blotted and washed. The tissues were blocked with 10% normal goat serum b and blotted. The samples were incubated with rabbit polyclonal anti-cpv (vp1/vp2) antibody (1 : 1,000) for 2 hours at 37 C and then blotted and washed. Prediluted biotinylated goat anti-rabbit secondary antibody (Zymed kit) was applied and incubated for 15 minutes at room temperature and then blotted and washed. The samples then were incubated with prediluted streptavidin peroxidase (Zymed kit), blotted, washed, visualized with 3-3 -diaminobenzidine tetrahydrochloride (DAB), and counterstained with hematoxylin. The samples were mounted with Permount c and examined by light microscopy. DNA Extraction Paraffin-embedded sections of cerebellum from each animal were used for DNA extraction as described previously. 11 Briefly, 2 4, 8- m sections were cut from each block with a microtome, ensuring that the blade was changed between blocks and that the blade holder also was cleaned with acetone. Blocks from animals with cerebellar hypoplasia were sectioned after those from control samples as a precaution against carry-over contamination. DNA was extracted from tissue sections with the QIAmp blood and tissue kit d by the manufacturer s instructions but with omission of the dewaxing step. Polymerase Chain Reaction Three PCR primer pairs were used to amplify parvoviral DNA; 2 pairs have been described previously. 11 These primer pairs amplify genes that code for structural proteins (VP1 and VP2) from both feline and canine parvovirus and do not discriminate between the 2 because the PCR primers anneal to conserved genomic regions. Primer pair 1 was designed to amplify a 215 base pair product from the VP2 gene. The sense primer corresponds to positions 3,724 3,744 (5 ACGTGGTGTAACTCAAATGG 3 ), and the antisense primer is complementary to positions 3,920 3,940 (5 ACCATGTTGTCTAC- CAAATGC 3 ). Primer pair 2 was designed to amplify a 193 base pair product from the VP1 gene. The sense primer corresponds to positions 2,320 2,339 (5 GGGTGTGTTAGTAAAGTGGG 3 ) and the antisense primer is complementary to positions 2,494 2,513 (5 CAGAGCGAAGATAAGCAGCG 3 ). Primer pair 3 was designed to amplify a 184 base pair product from the VP2 gene. The sense primer corresponds to positions 3,623 3,643 (5 CAAACAAATAGA- GCATTGGGC 3 ) and the antisense primer is complementary to positions 3,785 3,808 (5 GGTGCACTATAACCAACCTCAGC 3 ). A primer pair for the histone 3.3 (H 3.3) gene also was used to assess the integrity of the DNA extracted from the brains. These primers correspond to nucleotide positions 547 566 (5 CCACTGAACTTCT- GATTCGC 3 ) and 742 761 (5 GCGTGCTAGCTGGATGTCTT 3 ) and amplify a 214 base pair product. Reactions were carried out in a 50- L final volume containing Promega buffer diluted to 1, 1.5 mm MgCl 2, 200 mm dntp, 250 ng of each primer, and 0.25 units of Promega Taq polymerase. e Positive PCR controls included parvovirus genomic DNA extracted from parvoviral cultures, as well as genomic DNA extracted from the fresh cerebellum of a kitten with cerebellar hypoplasia. Two types of negative PCR controls were used, including sterile water extracted in parallel with the paraffin samples and a routine negative control with water as the PCR template. Amplifications were performed for 40 cycles with the use of the following parameters: denaturation for 5 minutes at 95 C, annealing for 1 minute at 55 C, extension for 1 minute at 72 C, followed by 39 cycles of 1 minute at 95 C, 1 minute at 55 C, and 1 minute at 72 C. Eight microliters of the 50- L PCR reaction product were analyzed by electrophoresis on a 1.5% agarose gel, and the bands were visualized under ultraviolet light after incorporation of ethidium bromide. Any resultant positive PCR products were purified. f Bidirectional sequencing of the purified PCR products was carried out with the forward and reverse primers from the original PCR reaction with Prism dye-terminator chemistry and a 3700 DNA sequencer as described by the manufacturer s protocol. g After 25 cycles of denaturation, annealing, and extension, the dye-terminated transcripts were ethanol precipitated, washed with 70% ethanol, and air dried before analysis. A homology analysis of the generated sequences was performed. Results Clinical signs of dogs (cases 1 6, 9 11) with cerebellar disease were noted when the animals 1st began to walk between 2 and 4 weeks of age; all manifested cerebellar ataxia with hypermetric movements of the limbs. History, onset, and progression of clinical signs are unknown for cases 7 and 8. Gross and histopathologic findings in dogs 1 and 2 have been reported previously. 8 Briefly, each had gross hypoplasia of the cerebellar vermis, and the flocculus was reduced in size in dog 2. Histopathologically, cerebellar cortex was relatively normal with the exception of mild central chromatolysis of Purkinje cells. Dog 3 was 1 of 9 Doberman Pinscher puppies in a litter. Several in the litter died; 3 of the pups manifested cerebellar-vestibular ataxia from birth. Postmortem examination disclosed a grossly atrophic cerebellum (7% of the total weight of the brain; normal, 10 12% of total brain weight; delahunta, Summers, and Kornegay, personal communications) with an underdeveloped vermis and small paramedian lobules that were situated transversely across the midline in the place of the vermis. The cerebellar hemispheres were asymmetric. On microscopic examination, the cerebellar cortex that was present was normal. Postmortem examination on dog 4 identified extensive malformation of the cerebellum which was also symmetrically smaller than normal. The vermis was incompletely developed and dysplastic. In addition, the colliculi of the midbrain were fused and there was no corpus callosum. The mesencephalic aqueduct was hypoplastic, and the lateral ventricles were dilated moderately. Histopathologically, the vermis was of normal microscopic architecture. The mass of the cerebellar tissue at the site of the nodulus consisted
540 Schatzberg et al of granular, molecular, and Purkinje cells and white matter layers, but these were abnormal in their arrangement. Islands of granule cells were present in the molecular layer, and some areas of Purkinje cells were depleted of granule cells. No formed folia were observed, and the outer contours were rounded and smooth. Postmortem examination of dog 5 identified a complete absence of the caudal portion of the cerebellar vermis. The rostral vermis and the hemispheres were slightly reduced in overall size. On histopathologic examination, the vermalcentral part of the cerebellar medulla was absent. No fastigial nuclei were present. The interposital nuclei were atrophic. No abnormalities were detected in the lateral cerebellar nuclei. Postmortem examination of dog 6 disclosed gross hypoplasia of the cerebellar vermis and the flocculus. The remainder of the cerebellum was normal on gross examination and histopathologically the cerebellar cortex was normal. Postmortem examination of dogs 7 and 8 (mixed breed littermates) identified mild diffuse atrophy of the cerebellum on gross examination. Histopathology on dogs 7 and 8 showed active, severe degeneration and depletion of the Purkinje cells in the cerebellar cortex consistent with a cortical cerebellar abiotrophy. No abnormalities were found on analysis of cerebrospinal fluid (CSF) performed on dogs 9 and 10 (Blue Tick Coonhound littermates). Sagittal magnetic imaging on the brain of dog 9 showed a mild increase in CSF between dorsal cerebellar folia compared to the ventral cerebellar folia, suggestive of folial atrophy and cerebellar hypoplasia (data not shown). On gross postmortem examination, the cerebella of dogs 9 and 10 were diffusely hypoplastic with no vermal defects or other lesions identified. The cerebella comprised 7.3 and 7.0% of total brain weight in dogs 9 and 10, respectively. Both had identical histopathologic findings comprising generalized atrophy of the cerebellum, which occasionally resulted in widening of the subarachnoid space in a sulcus (Fig 1). Both had normal Purkinje cell numbers but moderate reduction in granule cell neurons (Fig 1). The molecular layer was mildly thinned and hypercellular, with the increased cells thought to be either germinal cells that had failed to migrate into the granule cell zone or interneurons that were more concentrated than normal as a consequence of loss of granule axons. Clinical signs in the 8 cats were static except in cat 13, which deteriorated 2 days before euthanasia. Results of CSF evaluation in cat 13 were consistent with meningoencephalitis, with 9 white blood cells/ L (normal, 5 cells/ L) and 110 mg/dl protein (normal, 25 mg/dl). On gross postmortem examination, a diffusely hypoplastic cerebellum was observed, and histopathologic evaluation identified hypoplasia of the granular and molecular layers and mild hydrocephalus. The 7 other cats (cases 12, 14 19) had diffusely hypoplastic cerebella on gross examination, and histopathologic evaluation in all cats identified hypoplasia of the cerebellar granule layer 5 with sporadic Purkinje cell loss, consistent with a diagnosis of in utero infection with parvovirus. PCR amplification of the histone (H3.3) gene yielded the expected 214 base pair amplicon on all samples, confirming the integrity of the DNA extracted from paraffin-embedded cerebella (Fig 2b: lanes 2 6). Amplification of the DNA extracted from the hypoplastic cerebella of all cats (cases 12 19) produced PCR products of the expected size with the use of all 3 primer pairs for parvovirus (Fig 2a: lane 2). Amplification of the DNA extracted from the hypoplastic cerebella from the 2 Blue Tick Coonhound littermates (cases 9 and 10) yielded PCR products of the expected size with the use of all 3 primer pairs for parvovirus (Fig 2a: lanes 3 and 5). PCR amplification of the 6 dogs with cerebellar vermal defects (cases 1 6), the 2 littermates with cerebellar abiotrophy (cases 7 and 8), 1 dog with hydranencephaly (case 11), and 15 control animals yielded no products of the expected size with the use of any of the 3 parvoviral primer pairs (Fig 2a: lanes 4 and 6, primer pair 3 on dogs 1 and 7; data not shown for others). PCR negative control reactions for both sterile water extractions and water template with the use of both parvovirus and histone primers produced no amplicon in all cases (Fig 2a,b: lanes 7 and 8). Sequence analysis of PCR products amplified from the 8 felines (12 19) and both canines (9 and 10) of cerebellar hypoplasia confirmed that each amplicon was 98% conserved with canine parvovirus (data not shown). Immunohistochemical staining of deparaffinized cerebellar tissue from feline case 18 and both Blue Tick Coonhound littermates (cases 9 and 10) was negative for parvovirus. Positive controls, with the use of deparaffinized gastrointestinal samples from a previously confirmed canine parvovirus case, showed strong immunoreactivity within intestinal crypt epithelium. Discussion The most common cause of cerebellar hypoplasia in neonatal animals is an intrauterine viral infection during early or midgestation. A number of different viruses are capable of causing this malformation, including herpes simplex virus, 13 bluetongue virus, 14 Chuzan virus, 15 Akabane virus, 16 rift valley fever virus, 17 Cache Valley and Aino viruses, 4,17 the pestiviruses of bovine virus diarrhea (BVD) and hog cholera, 1,2 the flavivirus of Wesselbron disease, 17 Rubella, and the parvovirus of feline panleukopenia. 18 Cerebellar hypoplasia has been recognized in cats as early as 1888. 19 Clinical signs include cerebellar-vestibular ataxia with loss of balance, basewide stance, and a spastichypermetric gait. For many years, it was believed that this disease was genetic in origin, but in 1967, Johnson, Margolis, and Kilham identified feline panleukopenia virus as the definitive cause. 20 The pathogenesis is now well established. The virus has a predilection for cells with active DNA synthesis, such as those of the external germinal layer of the cerebellum, which remains active until 10 weeks of age in dogs and cats. The virus destroys this cell layer, which normally gives rise to the cerebellar granule cells and consequently causes hypoplasia of the granule layer (granuloprival hypoplasia), as well as having a separate cytopathic effect on Purkinje cells. 21 The end result often is a remarkable, diffuse hypoplasia or atrophy of the cerebellum. Furthermore, feline parvovirus can persist in mitotically active cells such as endothelial cells and macrophages, and so can cause severe dysplasia of the cerebellum as it
Cerebellar Parvovirus 541 Fig 1. (a) Two folia are shown from a normal cerebellum from a 6-month-old dog. H&E, 4. Bar 313 m. (b) Enlargement of 1 folium showing normal arrangement of molecular, Purkinje, and granule layer. Notice the greater width of the granule layer compared to panel d. H&E, 10. Barr 125 m. (c) Case 9: Several folia are seen about a central sulcus, which is dilated after cerebellar hypoplasia. M, meninges. H&E, 4. Bar 337 m. (d) Case 9: Enlargement of 1 folium reveals a narrowed granule cell layer and a hypercellular molecular layer. Purkinje cells and folial white matter appear intact. H&E, 10. Bar 135 m. continues to develop later in gestation and during the immediate postnatal period. 5,22 In 1984, Johnson and Castro isolated canine parvovirus from the brain of a 7.5-week-old Dalmatian dog with enteric disease and severe necrotizing vasculitis and encephalomalacia. 23 The cerebrum had multiple areas of leukomalacia with disseminated endothelial hypertrophy in the cerebellum and medulla, but no evidence of cerebellar hypoplasia. Canine herpes virus is another cause of extensive inflammation in multiple body systems, including the central nervous system. 24 Acute encephalitis can be seen with herpes virus infection, and if a puppy survives, it might have a cerebellar ataxia from the residual affects of cerebellar inflammation. No viral etiology has been identified as a direct cause of cerebellar hypoplasia in the dog to date. Cerebellar hypoplasia appears to be uncommon in the dog, but the precedent for in utero viral destruction of the cerebellum in other species makes a viral etiology likely.
542 Schatzberg et al Fig 2. (a) PCR amplification for parvoviral DNA with the use of primer pair 3 (184 base pair product). The lower bands represent primer dimer of 100 bp. Lane 1: 100-bp DNA ladder; arrow denotes 500 bp. Lane 2: case 18: feline cerebellar hypoplasia, positive (data not shown for cats 12 17, 19; data not shown for primer pairs 1 and 2). Lane 3: case 9: Blue Tick Coonhound 1 with cerebellar hypoplasia, positive (data not shown for primer pairs 1 and 2). Lane 4: case 1: canine vermal defect, negative. Lane 5: case 9: Blue Tick Coonhound 2 with cerebellar hypoplasia, positive (data not shown for primer pairs 1 and 2). Lane 6: case 7: canine cerebellar cortical abiotrophy, negative. Lane 7: DNA extraction negative control, negative. Lane 8: PCR negative control water template, negative. (b) PCR amplification for histone (H3.3) DNA (214 base pair product). Lanes 1 8: same as panel a. The positive PCR for canine parvovirus in the 2 Blue Tick Coonhound littermates, coupled with histopathologic studies identifying granuloprival hypoplasia, strongly implicates a parvoviral etiology. Careful study of the records of these dogs indicates that the dam was vaccinated for parvovirus early in pregnancy. One possibility is that an incompletely attenuated vaccine virus was responsible for an in utero infection and cerebellar destruction in these dogs. However, wild-type parvovirus infection as the cause for cerebellar hypoplasia in these 2 dogs cannot be excluded. Canine parvovirus isolates have been classified into 2 major subtypes, CPV type-2 and CPV type-2a, which are distinguished by antigenic, biological, and genetic criteria. 25 The original CPV type-2 strain was present in nature only between 1978 and approximately 1980, when it was replaced by the CPV type-2a strain. Around 1984, a further minor antigenic variant of the CPV type-2a strain was recognized and was termed CPV type-2b. 25 The sequence of the virus from the brains of dogs 9 and 10 obtained from the PCR products covered the region between bases 2,321 and 2,514 and between bases 3,623 and 3,910, respectively, of the complete CPV genome. 26 From these sequences, it was clear that the virus was a CPV isolate, and the sequences at positions 3,285 (G determining Gly at VP2 residue 300) and 3,689 (T determining Tyr at VP2 residue 305) indicated that the virus was either a CPV type-2a or
Cerebellar Parvovirus 543 CPV type-2b virus, whereas the sequence at position 3,675 (T determining Ser at VP2 residue 297) is characteristic of a virus infecting dogs in the United States between 1980 and 1995. 26 This finding indicates that the virus that infected the Coonhound dogs was a wild type virus, because vaccines being used in the early 1990s, when the Coonhound pups were affected, contained the original (1978 1980) CPV type-2 strain of virus. Interestingly, immunohistochemistry (IHC) for parvoviral antigen was negative on both littermates as well as on a 6-week-old kitten with cerebellar hypoplasia. However, IHC also failed to demonstrate the presence of parvovirus in over 15 hypoplastic feline cerebella after 2.5 weeks of age, 21 suggesting that as the infection becomes chronic, reduced viral load occurs with minimal translation of viral proteins. Negative IHC results in cases 9, 10, and 18, therefore, are consistent with previous immunohistochemical studies and suggest that PCR is a more sensitive diagnostic technique for the detection of parvovirus in the brains of cats and dogs with cerebellar hypoplasia. A much more common cause of inherited cerebellar disease in the dog is cortical abiotrophy, for which a recessive pattern of inheritance has been suspected in many breeds. 5,10 Cerebellar vermal malformations, or Dandy Walker-like defects, also occur sporadically in the dog. 8,9 In humans, Dandy Walker syndrome is considered to be a primary parenchymal, midline, developmental defect of genetic origin. 27,28 Our inability to amplify parvoviral DNA from either of the dogs with vermal defects or from the littermates with cerebellar abiotrophy is consistent with a genetic pathogenesis. Having amplified parvoviral DNA from 2 kittens with cerebellar hypoplasia and hydranencephaly, 11 we also studied 1 Labrador Retriever with congenital hydranencephaly. The negative PCR for parvovirus in the Labrador Retriever is consistent with nonviral causes of hydranencephaly. 12 The negative results from dogs with presumed genetic disorders and our inability to amplify parvovirus from any of the 15 control animals suggest that parvoviral DNA cannot be randomly amplified from canine and feline brains. The positive results for the kittens with cerebellar hypoplasia are consistent with the previously established viral pathogenesis. The presence of parvoviral DNA in the Blue Tick Coonhound littermates represents a novel etiology for cerebellar hypoplasia in the dog. Although the canine and feline parvoviruses are over 99% identical in DNA sequence, they differ in several biological properties. 26 It therefore is not known whether the difference in apparent frequency of parvovirus-induced cerebellar hypoplasia is due to differences in the virus or the host or to differences in both. Although canine cerebellar hypoplasia seems to be uncommon, canine parvovirus infection should be a differential diagnosis for puppies with congenital cerebellar disease. The results obtained from animals with cerebellar hypoplasia confirm that parvoviral DNA can be amplified from formalin-fixed, paraffin-embedded brains of both dogs and cats. The size of PCR products that can be amplified from paraffin blocks depends on the primers themselves, the type of tissue, the type of fixative, and how long the tissue remains in the fixative. 29,30 Because formalin can crosslink DNA, when using PCR on formalin-fixed tissues, primer pairs ideally should be designed to produce products of approximately 200 base pairs in size. 29,30 In order to maximize the generation of visible PCR products, the primer pairs in this study were chosen to be close to 200 base pairs in length and each was used for 40 rounds of amplification. To eliminate DNA degradation as a cause for false negative data, we optimized the amplification of a segment of parvorviral DNA by utilizing 3 independent parvoviral primer pairs and also amplified a segment of the canine and feline histone (H3.3) gene as a positive control. This methodology now has proven utility in testing for parvovirus in feline and canine cerebellar hypoplasia. Likewise, this test might prove useful in meningoencephalitis of unknown etiology in the dog and cat, because parvovirus B19 DNA has been detected by PCR in recent reports of acute meningoencephalitis in humans. 31 Footnote a ProbeOn slides, Biomeda, Foster City, CA b Zymed Histostain SP Rabbit kit, Zymed Laboratories Inc, South San Francisco, CA c Permount, Fisher Scientific, Suwanee, GA d QIAmp blood and tissue kit, Qiagen, Inc, Chatsworth, CA e Taq polymerase, Promega Corporation, Madison, WI f Strataprep PCR purification kit, Stratagene, Cedar Creek, TX g 3700 DNA sequencer, Applied Biosystems, Inc, Foster City, CA Acknowledgments The work was done primarily at Cornell University Hospital for Animals supported by an in-house research grant. Preliminary data was obtained with support from North Carolina State University, College of Veterinary Medicine (SJS and NJHS). References 1. Johnson KP, Ferguson LC, Byington DP, Redman DR. Multiple fetal malformations due to persistent viral infection. I. Abortion, intrauterine death, and gross abnormalities in fetal swine infected with hog cholera vaccine virus. Lab Invest 1974;30:608 617. 2. Barlow RM. Morphogenesis of hydranencephaly and other intracranial malformations in progeny of pregnant ewes infected with pestiviruses. J Comp Pathol 1980;90:87 98. 3. MacLachlan NJ, Osburn BI, Ghalib HW, Stott JL. Bluetongue virus-induced encephalopathy in fetal cattle. Vet Pathol 1985;22:415 417. 4. Kitano Y, Ohzono H, Yasuda N, Shimizu T. Hydranencephaly, cerebellar hypoplasia, and myopathy in chick embryos infected with aino virus. Vet Pathol 1996;33:672 681. 5. Summers B, Cummings J, delahunta A. Malformations of the Central Nervous System. Veterinary Neuropathology. St Louis, MO: Mosby; 1994:82 86. 6. Dow R. Partial agenesis of the cerebellum in dogs. J Comp Neurol 1940:569 586. 7. Pass DA, Howell JM, Thompson RR. Cerebellar malformation in two dogs and a sheep. Vet Pathol 1981;18:405 407. 8. Kornegay JN. Cerebellar vermian hypoplasia in dogs. Vet Pathol 1986;23:374 379. 9. Schmid V, Lang J, Wolf M. Dandy-Walker-like syndrome in four dogs: Cisternography as a diagnostic aid. JAHA 1992;28:355 360. 10. March P. Degenerative brain disease. In: Bagley R, ed. Intra-
544 Schatzberg et al cranial Disease, Vet Clinics of North America. Philadelphia, PA: WB Saunders; 1996:945 971. 11. Sharp NJ, Davis BJ, Guy JS, et al. Hydranencephaly and cerebellar hypoplasia in two kittens attributed to intrauterine parvovirus infection. J Comp Pathol 1999;121:39 53. 12. Barone G, delahunta A, Sandler J. An unusual neurological disorder in the Labrador retriever. J Vet Intern Med 2000;14:315 318. 13. Christie JD, Rakusan TA, Martinez MA, et al. Hydranencephaly caused by congenital infection with herpes simplex virus. Pediatr Infect Dis 1986;5:473 478. 14. Richards WP, Cordy DR. Bluetongue virus infection: Pathologic responses of nervous systems in sheep and mice. Science 1967;156: 530 531. 15. Muira Y, Kubo M, Goto Y, et al. Hydranencephaly-cerebellar hypoplasia in a newborn calf after infection of its dam with Chuzan virus. Jpn J Vet Sci 1996;52:689 694. 16. Chung SI, Livingston CW Jr, Edwards JF, et al. Congenital malformations in sheep resulting from in utero inoculation of Cache Valley virus. Am J Vet Res 1990;51:1645 1648. 17. Coetzer JA, Barnard BJ. Hydrops amnii in sheep associated with hydranencephaly and arthrogryposis with wesselsbron disease and rift valley fever viruses as aetiological agents. Onderstepoort J Vet Res 1977;44:119 126. 18. Kilham L, Margolis G. Viral etiology of spontaneous ataxia of cats. Am J Pathol 1966;48:991 1011. 19. Herringham WP, Andrews FN. Two cases of cerebellar disease in cats with staggering. St Bartholomew s Hosp Rep (Lond) 1888;24: 241 248. 20. Johnson RH, Margolis G, Kilham L. Identity of feline ataxia virus with feline panleucopenia virus. Nature 1967;214:175 177. 21. Csiza C, Scott F, Gillespie J, et al. Feline viruses XIV. Transplacental infections in spontaneous panleukopenia of cats. Cornell Vet 1971;61:423 439. 22. Margolis G, Kilham L. Parvovirus infections, vascular endothelium, and hemorrhagic encephalopathy. Lab Invest 1970;22:478 488. 23. Johnson BJ, Castro AE. Isolation of canine parvovirus from a dog brain with severe necrotizing vasculitis and encephalomalacia. J Am Vet Med Assoc 1984;184:1398 1399. 24. Percy DH, Carmichael LE, Albert DM, et al. Lesions in puppies surviving infection with canine herpesvirus. Vet Pathol 1971;8:37 53. 25. Parrish CR, Aquadro CF, Strassheim ML, et al. Rapid antigenictype replacement and DNA sequence evolution of canine parvovirus. J Virol 1991;65:6544 6552. 26. Parrish CR. Mapping specific functions in the capsid structure of canine parvovirus and feline panleukopenia virus using infectious plasmid clones. Virology 1991;183:195 205. 27. Gilbert-Barness E, Debich-Spicer D, Cohen MM Jr, Opitz JM. Evidence for the midline hypothesis in associated defects of laterality formation and multiple midline anomalies. Am J Med Genet 2001;101:382 387. 28. Fan YS, Siu VM. Molecular cytogenetic characterization of a derivative chromosome 8 with an inverted duplication of 8p21.3 p23.3 and a rearranged duplication of 8q24.13 qter. Am J Med Genet 2001;102:266 271. 29. Greer CE, Peterson SL, Kiviat NB, Manos MM. PCR amplification from paraffin-embedded tissues. Effects of fixative and fixation time. Am J Clin Pathol 1991;95:117 124. 30. Karlsen F, Kalantari M, Chitemerere M, et al. Modifications of human and viral deoxyribonucleic acid by formaldehyde fixation. Lab Invest 1994;71:604 611. 31. Barah F, Vallely PJ, Chiswick ML, et al. Association of human parvovirus B19 infection with acute meningoencephalitis. Lancet 2001;358:729 730.