• Open Access

Mutation in β1-Tubulin Correlates with Macrothrombocytopenia in Cavalier King Charles Spaniels


Corresponding author: Mary K. Boudreaux, Department of Pathobiology, Auburn University, AL 36849-5519; e-mail: boudrmk@auburn.edu.


Background: Cavalier King Charles Spaniels (CKCS) have a high prevalence of inherited macrothrombocytopenia. The purpose of this study was to determine if a mutation in β1-tubulin correlated with presumptive inherited macrothrombocytopenia.

Hypothesis: A mutation in β1-tubulin results in synthesis of an altered β1-tubulin monomer. α-β tubulin dimers within microtubule protofilaments are unstable, resulting in altered megakaryocyte proplatelet formation.

Animals: Blood samples were obtained from CKCS and non-CKCS dogs.

Methods: DNA was used in polymerase chain reaction (PCR) assays to evaluate β1-tubulin. Platelet numbers and mean platelet volume (MPV) were evaluated for a correlation with the presence or absence of a mutation identified in β1-tubulin. Platelets obtained from homozygous, heterozygous, and clear CKCS were further evaluated using electron microscopy and immunofluorescence.

Results: A mutation in the gene encoding β1-tubulin correlated with macrothrombocytopenia in CKCS. Electron microscopy and immunofluorescence studies suggest that platelet microtubules are present but most likely are unstable and decreased in number.

Conclusions and Clinical Importance: The macrothrombocytopenia of CKCS correlated with a mutation in β1-tubulin. α–β tubulin dimers within protofilaments most likely are unstable, leading to altered proplatelet formation by megakaryocytes. This information will aid in distinguishing inherited from acquired thrombocytopenia. It also provides insight into the mechanism of platelet production by megakaryocytes, and also may prove useful in understanding heart-related changes in macrothrombocytopenic CKCS with concurrent mitral valve regurgitation.

Cavalier King Charles Spaniels (CKCS) have a high prevalence of macrothrombocytopenia that is inherited as an autosomal trait.1–4 Affected dogs do not have a bleeding diathesis, but the existence of low platelet numbers can be confused with and must be distinguished from acquired causes of thrombocytopenia, including thrombocytopenia secondary to infectious agents, consumption, medications, immune-mediated causes, and others. Unfortunately, CKCS have received inappropriate treatment with antibiotics, corticosteroids, or other medications because of confusion or lack of awareness of this disorder.5 A molecular assay to confirm an inherited cause for macrothrombocytopenia would help veterinarians distinguish inherited from acquired macrothrombocytopenia and assist breeders in identifying carriers of the mutation. Identification of the causative mutation also would provide information on the mechanism of platelet production by megakaryocytes and provide clues to the severity of heart changes in CKCS with concurrent mitral valve regurgitation. The molecular basis for macrothrombocytopenia in CKCS has not been reported. Recently, a mutation in the β1-tubulin gene was noted at coding nucleotide 745 (c.745) by one of the authors (Boudreaux, unpublished observation) in several CKCS with macrothrombocytopenia. This mutation is predicted to change the encoded amino acid in location 249 from aspartic acid to asparagine (D249N). The purpose of this study was to determine the prevalence of the mutation and whether there was a correlation of the identified mutation with macrothrombocytopenia in CKCS.

Materials and Methods

This study was approved by the Auburn University (AU) Institutional Animal Care and Use Committee. Client consent forms were completed by CKCS owners when blood samples were collected for the purpose of this study alone. Samples used in this study were primarily obtained from surplus EDTA blood samples submitted to the AU College of Veterinary Medicine Clinical Pathology laboratory for CBC determinations. Platelet number and mean platelet volume (MPV) were determined for all samples obtained at AU.a The presence or absence of thrombocytopenia was verified for all samples obtained at AU by evaluating blood smears. Platelet size was evaluated by recording the percentage of platelets >25% the diameter of an erythrocyte in × 10 and × 400 magnification fields. Samples identified by both methods of evaluation were included in the study. CKCS samples also were kindly provided by North Carolina State University, Texas A&M University, University College Dublin's Veterinary Hospital, and the University of Florida as well as by private practitioners. One hundred samples were obtained from CKCS and 52 samples were obtained from non-CKCS. The ages of the CKCS were not available for all dogs studied, but for those in which age was known the dogs ranged from 4 months to >10 years of age. Samples were obtained from 48 males and 52 females. Platelet counts and MPV were not available for all of the CKCS samples evaluated by polymerase chain reaction (PCR). All of the non-CKCS samples were obtained at AU, and platelet counts and MPV were available for all of those samples. Samples from all dogs were evaluated regardless of platelet number or the presence of disorders that may have resulted in an acquired thrombocytopenia.


Genomic DNA was harvestedb from EDTA-anticoagulated blood or from blood smears obtained from CKCS and non-CKCS dogs. Primers were designed to ensure assessment of all coding areas and exon–intron splice sites in initial studies. Primers were designed using sequence obtained from the canine genome available on GenBank. The correct area of the canine genome was located by comparing predicted canine β1-tubulin cDNA sequence available on GenBank (gi:73992646). Once the mutation at nucleotide c.745 was documented in several CKCS with macrothrombocytopenia, primers were redesigned to flank this area, which became the focus of this study. PCR products were subjected to electrophoresis on agarose gels. Target bands were harvestedc for sequencingd and DNA sequence results were printed and evaluated.

Immunofluorescence (IF)

Blood (10 mL total volume) was collected from dogs that were clear (n = 2), heterozygous (n=1), and homozygous (n=2) for the c.745 mutation into 3.8% citrate at a ratio of 9 : 1. Platelet-rich plasma was prepared using differential centrifugation in a swinging bucket rotor. An aliquot of PRP was transferred to a microcentrifuge tube and 1 μL of PGE1e (1 mg/mL) was added per milliliter of PRP. Samples were gently mixed and incubated in a warm water bath for 15 minutes, with gentle periodic mixing. PRP samples then were diluted 1 : 1 with Tangen HEPES bufferf and diluted samples used for cytosping preparations onto glass slides.h Cytospin preparations were fixed for 30 minutes in methanol and then washed (3 min/wash) twice with water and 3 times with phosphate-buffered saline (PBS). After the final wash, the slides were gently dried and placed in humidity chambers. Primary antibodyi to β1-tubulin at a dilution of 1 : 2000 in blocking bufferj was pipetted onto the slides (for negative controls, blocking buffer without primary antibody was used) and the slides were incubated for 1 hour. At the end of 1 hour, slides were washed 3 times, 3 minutes each, in PBS. After washing, the slides were placed back in humidity chambers and secondary antibodyk at a dilution of 1 : 1000 was pipetted onto the slides. Slides were incubated for 1 hour in the dark and then washed 3 times in PBS. After the last wash, slides were gently dried and mountedl with coverslips and the coverslips were sealed. Slides were viewed using a fluorescent microscope.m

Electron Microscopy

Platelet-rich plasma was prepared from 1 dog in each genetic category (clear, heterozygous, and homozygous affected) as described above. One milliliter of PRP from each dog was transferred to microcentrifuge tubes and PGE1 added. Then 0.75 mL of glutaraldehyde (0.1%) in White's salinen was added to the microcentrifuge tubes, and the tubes were gently mixed and incubated for 15 minutes at 37 °C with gentle periodic mixing during incubation.6 After 15 minutes, platelets were sedimented in a microcentrifuge and the supernatant was discarded and replaced with 0.75 mL 3% glutaraldehyde in White's salineo and allowed to fix for 1 hour at room temperature (RT). The supernatant was discarded and the pellet was washed with the same buffer. The cells were then postfixed in 1% osmium tetroxide in White's saline at RT for 1 hour and dehydrated in a graded series of ethanols. Propylene oxide was used as a transient solvent. Samples were embedded in Epon 812 resin.p Ultrathin sections were stained with uranyl acetate and lead citrate and examined with a Philips 301 electron microscope.q

Statistical Analysis

The general linear model (GLM) for analysis of variance (ANOVA) was used for comparisons of platelet counts and MPV by sequence. Specific comparisons were made using contrasts. Statistical Analysis System (SAS) softwarer was used for these analyses. Distribution of the mutation data comparing U.S. with Dublin samples was performed by the χ2 test with a software package.s


PCR assays were performed on DNA samples isolated from the blood of 100 CKCS and 52 non-CKCS. Of the 100 CKCS samples evaluated by PCR, 19 (19%) matched the canine genome at the mutation site (c.745 G), whereas 48 (48%) were heterozygous (c.745 G/A) and 33 (33%) were homozygous (c.745 A) for the mutation. When the 40 CKCS samples from Dublin were evaluated as a separate group, 14 (35%) matched the canine genome, 21 (52.5%) were heterozygous, and 5 (12.5%) were homozygous affected. These findings contrasted with those found for the 60 CKCS samples obtained in the United States in which 5 (8%) were found to match the canine genome, 27 (45%) were heterozygous, and 28 (47%) were homozygous affected (Table 1). Of the 52 non-CKCS samples evaluated by PCR, 51 matched the canine genome at the mutation site and 1 (Labrador Retriever) was heterozygous at this site.

Table 1.   Comparison of Dublin and U.S. CKCS genotypes for the β-1 tubulin mutation at c. 745.
DNA sequencen = 40n = 60
DublinM/FUnited StatesM/F
  1. P < .0001 for all quadrants by χ2 analysis.

  2. GAC, homozygous clear; G/AAC, heterozygous; AAC, homozygous affected; M/F, male/female distribution; CKCS, Cavalier King Charles Spaniels.

GAC14 (35%)9/55 (8%)0/5
G/AAC21 (52.5%)9/1227 (45%)14/13
AAC5 (12.5%)3/228 (47%)13/15

Platelet counts were available from 89 CKCS and from all 52 non-CKCS. MPVs were available from 81 CKCS and from all 52 non-CKCS.

CKCS platelet counts ranged from 34,000 to 666,000/μL whereas non-CKCS platelet counts ranged from 24,000 to 698,000/μL. CKCS MPV ranged from 7.8 to 35.4 fL and non-CKCS MPV ranged from 7.9 to 23.6 fL.

CKCS samples were divided into groups based on the presence or absence of the β1-tubulin gene mutation at c.745, and platelet numbers were reevaluated based on these groups (Table 2). Platelet numbers were significantly different among groups, with low platelet counts (generally <100,000/μL) correlating with a homozygous state for the mutation, intermediate platelet counts (approximately 200,000/μL) correlating with a heterozygous state for the mutation, and higher platelet numbers (generally >250,000/μL) correlating with being clear of the mutation. Because CKCS with acquired thrombocytopenia were included in the calculations, low platelet counts did not always correlate with the presence of the mutation. Four CKCS that were homozygous affected for the c.745 mutation had platelet counts >100,000/μL, and their platelet numbers ranged from 122,000 to 171,000/μL. These samples were not obtained at AU and platelet numbers could not be evaluated by other means such as smear evaluation. The 1 non-CKCS dog that was heterozygous for the c.745 mutation had platelet numbers in an intermediate range comparable to that observed in the heterozygous CKCS dogs. MPVs also were significantly different when comparing the 3 groups to one another (Table 2).

Table 2.   Platelet numbers and MPV of CKCS and non-CKCS dogs grouped based on the presence or absence of a missense mutation in the β1-tubulin gene at c.745.
CKCS (n)Platelet count
(× 1000/μL)
Non-CKCS (n)Platelet count
(× 1000/μL)
GAC18366 ± 14851346 ± 160
G/AAC45196 ± 641130
AAC2674 ± 360NA
CKCS (n)MPV (fl)Non-CKCS (n)MPV (fl)
  1. Values are mean ± SD.

  2. Means between GAC, G/AAC, and AAC groups were significantly different from each other in CKCS for platelet number and MPV (P≤ .0002).

  3. NA, not applicable; GAC, homozygous clear; G/AAC, heterozygous; AAC, homozygous affected; CKCS, Cavalier King Charles Spaniels; MPV, mean platelet volume.

GAC1612 ± 2.25112 ± 3
G/AAC4316.5 ± 3.2117.5
AAC2228 ± 5.70NA

Microtubules were intact and readily seen as continuous lariat-like structures in the outer cytoplasmic boundary in nonactivated platelet samples obtained from dogs clear of the β1-tubulin mutation when viewed using fluorescent microscopy (Fig 1A). In contrast, platelets obtained from dogs that were either heterozygous or homozygous for the mutation had microtubule structures that appeared granular and discontinuous (Fig 1B,1C). The location of microtubule staining was similar to that seen in normal dog samples, but the lariat-like structures were indistinct.

Figure 1.

 Cytospin preparations of platelets obtained from clear (A), heterozygous (B), and affected (C) Cavalier King Charles Spaniels (CKCS) for the c.745 mutation in β1-tubulin. Platelet preparations were incubated with a mouse monoclonal antibody to β1-tubulin followed by incubation with a goat anti-mouse antibody labeled with Alexa Fluor 488. Samples obtained from a clear CKCS had distinct microtubule coils in the periphery of the cells. Microtubule coils were present but not distinct in samples obtained from heterozygous or affected CKCS.

A marginal band of microtubules containing 10–30 coils (average, 17.3 coils/platelet) was readily apparent along the long axis of platelets when platelet samples obtained from a normal dog were evaluated by electron microscopy (Fig 2A). Cross-sectional profiles of microtubules were present in approximately 10% of the platelets. Microtubules were not readily seen in platelet samples obtained from a dog homozygous for the mutation (Fig 2C). Only 1 or 2 microtubule profiles were seen in the affected platelets. After searching extensively, 1 platelet was discovered with 8, abnormally assembled, short microtubule-type profiles. In platelets obtained from a heterozygous CKCS, the average number of coils per platelet was 7.6 and ranged from 1 to 16. Cross-sectional profiles of microtubules were observed in approximately 5% of the cells (Fig 2B).

Figure 2.

 Electron micrographs of platelet samples obtained from clear (A), heterozygous (B), and affected (C) Cavalier King Charles Spaniels (CKCS) for the c.745 mutation in β1-tubulin. Microtubules (black arrows point to longitudinal sections of microtubules; white arrows point to cross sections of microtubules) were readily seen in samples obtained from a clear dog and marginal bands contained an average of 17.3 coils/platelet. In contrast, microtubules were not readily seen in samples obtained from an affected dog. After searching extensively, 1 platelet was found with 8 abnormally assembled, short microtubule-type profiles. Samples obtained from a heterozygous CKCS were intermediate in terms of microtubule numbers and structure. Scalebar =100 nm.


CKCS have a high prevalence of asymptomatic inherited macrothrombocytopenia.1–4 In this study, a missense mutation (c.745 G>A) in the gene encoding β1-tubulin, predicted to result in the substitution of an asparagine for aspartic acid (D249N), correlated with the macrothrombocytopenia observed. The charge change resulting from this substitution is speculated to result in impaired microtubule assembly affecting proplatelet formation and platelet production by megakaryocytes. Interestingly, the distribution of the mutation in CKCS from Dublin differed from that observed in the United States with more dogs identified as homozygous clear and heterozygous and fewer dogs identified as homozygous affected in Dublin (Table 1). Whether this difference is related to sample size, breeding practices, or a reflection of the mutation being introduced into the Dublin population at a later time point is not known.

Microtubules consist of heterodimers composed of α and β monomers that associate head-to-tail to form protofilaments.7 Protofilaments in turn associate laterally in a left-handed helical fashion to form cylindrical and hollow microtubules. Microtubules perform a variety of functions in many cells and can be dynamic or stable. Examples of dynamic functions include involvement in cell mitosis and organelle movement, whereas examples of stable functions include providing support within cilia and flagella. There are many isoforms of α and β monomers that are encoded by separate genes. Although the isoforms are highly homologous, they do vary particularly at their carboxy-terminal domains, which may be important in determining microtubule stability for a particular function. Hydrolysis of GTP is required for microtubule assembly. α and β monomers bind GTP, but GTP is exchangeable only on β monomers. Microtubule stability is maintained by a cap of GTP bound at the tubulin ends.

Tubulin monomers are composed of 3 major domains. The amino-terminal residues (1–206) form alternating parallel β sheets and helices and are involved in GTP binding and hydrolysis. The center residues (207–384) are primarily involved in contacts between monomers that include both inter- and intraprotofilament interactions. The third domain is composed of 2 long helices and an interconnecting loop, and (with other residues) is most likely important for assembly of microtubules and interaction with microtubule-associated proteins (MAPs).8 Directed mutations in the β-tubulin gene mec-7 of Caenorhabditis elegans resulted in identification of distinct regions that were likely to be involved in microtubule assembly.9 These discrete regions included amino acids 171–188, 243–249, and 286–318. Specifically, a missense mutation (GAC to AAC) of the codon encoding amino acid 249 with the resultant change of aspartic acid to asparagine (D249N) was associated with loss of sensitivity to touch, most likely as a result of microtubule instability in touch receptor neurons. This amino acid is thought to be a part of an intraprotofilament binding site, and a change in charge at this site most likely interferes with microtubule assembly in C. elegans. Because this mutation is in the identical location as the one described in this report in macrothrombocytopenic CKCS, microtubule instability in affected CKCS megakaryocytes may lead to the altered thrombopoietic mechanism observed in these dogs. Evaluation of homozygous-affected CKCS megakaryocyte proplatelet formation in culture would provide valuable information to evaluate this hypothesis.

One β1-tubulin polymorphism (Q43P) has been described in people.10 The mutation was documented in 10.6% of the general population and in 24.2% of 33 unrelated people with inherited macrothrombocytopenia. Heterozygotes had spherocytic platelets and reduced β1-tubulin but normal platelet numbers. Platelet aggregation responses to thrombin receptor activation peptide (TRAP) and collagen and ATP release in response to low concentrations of ADP and collagen were impaired. Although platelet function differences were not found when comparing heterozygous men with heterozygous women, the mutation seemed to protect men but not women from cardiovascular disease. In another study of the Q43P polymorphism in humans, men were found to be protected from thrombotic disorders but were at increased risk for intracerebral hemorrhage.11 The mutation in β1-tubulin identified in this study in CKCS has not yet been reported in people with macrothrombocytopenia.

β1-tubulin is the isoform of β tubulin that is primarily expressed in megakaryocytes and platelets12 and is 1 of only 2 lineage-specific components involved in platelet release from megakaryocytes that has been identified. Knock-out models for NF-E2, an erythro-megakaryocytic transcription factor, lacked circulating platelets as a result of failure of production of proplatelets.13 Megakaryocytes of these mice completely lacked β1-tubulin. As a result of these studies, β1-tubulin knock-out models were developed.12β1-tubulin knock-out animals were fully developed and did not show overt signs of hemorrhage. Affected mice were thrombocytopenic, and circulating platelets had decreased numbers of microtubule coils. The marginal band of microtubules was not only decreased but also showed kinks and breaks when viewed by confocal immunofluorescence microscopy. Platelet size was not increased, but platelet shape was described as being spherical rather than elliptical. β2 and β5 isoforms of tubulin were increased in knock-out platelets, suggesting that upregulation of these 2 isoforms partially compensates for the loss of the β1 isoform. Upregulation of these isoforms apparently was not sufficient to compensate for loss of β1-tubulin. Although null mice did not show overt signs of hemorrhage, bleeding times were prolonged compared with wildtype mice. The prolonged bleeding time was thought to be disproportionate to the severity of thrombocytopenia observed, and it was speculated that the marginal band of microtubules plays a role in platelet function as well as in platelet formation. β1-tubulin null platelets also showed an attenuated response to thrombin activation in vitro.14

CKCS with inherited macrothrombocytopenia are not prone to hemorrhage and do not have prolonged bleeding times. Reports of platelet function in affected CKCS have been contradictory, with some reports indicating platelet hyporesponsiveness and others reporting platelet hyperresponsiveness.15,16 These discrepancies most likely are attributable to variations in methodology and in the CKCS population evaluated. For example, PFA-100 closure times (CT) were prolonged in CKCS dogs with mitral valve regurgitation suggesting platelet hyporesponsiveness.17 The prolonged CT values most likely were caused by enhanced cleavage of von Willebrand factor (vWF) by ADAMS13 (a disintegrin and metalloprotease with thrombospondin repeats), an enzyme that cleaves vWF under high shear conditions, and are not related to the inherited macrothrombocytopenia. Platelet aggregometry studies also can have variable results depending on the method used to isolate PRP. Platelet isolation techniques that fail to include all platelet populations in PRP samples can result in misleading findings (Boudreaux, personal observations). The lack of clinical bleeding in macrothrombocytopenic CKCS suggests that the microtubule defect does not have a major effect on platelet function. The maintenance of sufficient platelet mass is most likely responsible for the normal bleeding times observed in macrothrombocytopenic dogs. This finding is in contrast to many of the inherited macrothrombocytopenias documented in humans which usually are associated with bleeding tendencies and platelet hyporesponsiveness18 because of altered glycoproteins essential for normal platelet reactivity. In this report, electron microscopy and immunofluorescence findings suggested that affected CKCS platelets do have β1-tubulin; however, assembly of microtubules most likely is not efficient, leading to dysfunctional proplatelet production and platelet formation. CKCS also are known to develop mitral valve regurgitation. There has been speculation that common mutations are involved with these disorders, but Cowan et al were unable to document an association between the presence of macrothrombocytopenia and the presence of a heart murmur in a study of 69 CKCS. Interestingly, in a separate study of pressure overload-induced cardiac hypertrophy in cats, hypertrophied cardiocytes developed increased microtubule density as a result of marked upregulation of the transcription of genes encoding β1- and β2-tubulin isoforms.19 Increased microtubule density associated with increased microtubule polymerization has been linked to cardiocyte apoptosis in the pathogenesis of heart failure. Studies in rats and dogs documented that colchicine, a microtubule depolymerizing agent, significantly suppressed cardiocyte apoptosis and enhanced microtubule depolymerization and ventricular contractility, and thus was suggested as a potential treatment in preventing heart failure.20,21 Upregulation of mutated unstable β1-tubulin isoforms in response to pressure or volume overload in macrothrombocytopenic CKCS may play a role in alleviating the severity of heart-related clinical signs observed in some of these dogs. More studies are necessary to determine if altered β1-tubulin isoforms play a role in the progression of mitral valve regurgitation or other heart disorders.

In summary, a mutation in a gene encoding a megakaryocyte-specific protein, β1-tubulin, is associated with the macrothrombocytopenia observed in CKCS. A mutation in this gene correlates with the cell-specific effects seen (ie, macrothrombocytopenia as a result of altered platelet formation by megakaryocytes). Cell culture of megakaryocytes obtained from affected CKCS is necessary to provide direct evidence of altered proplatelet formation in these dogs and potentially would provide valuable insight into the mechanism of proplatelet formation and platelet release by megakaryocytes. Although the β1-isotype of β tubulin is considered to be megakaryocyte-specific, this isoform may be upregulated in other cell types such as cardiocytes, under certain pathologic conditions. This may be of particular relevance in macrothrombocytopenic CKCS with concurrent cardiac conditions. Although the mutation was not found in high prevalence in non-CKCS dogs, 1 Labrador Retriever was found to be heterozygous for the mutation. Future studies should include evaluation of the β1-tubulin gene for the presence of mutations in non-CKCS dogs with unexplained persistent macrothrombocytopenia.


aADVIA 120, Siemens Medical Solutions USA Inc, Malvern, PA

bQIAamp DNA Blood Mini Kit, Qiagen Inc, Valencia, CA

cQIAquck Gel Extraction Kit, Qiagen Inc

dABI 3100 Genetic Analyzer, Applied Biosystems, Foster City, CA

eSigma-Aldrich, St. Louis, MO

f10 mM HEPES, 145 mM NaCl, 5 mM KCl, 1 mM MgSO4, and pH 7.4, Sigma-Aldrich

gWescor Aerospray Hematology Slide Stainer/Cytocentrifuge Model 7120 Wescor Biomedical Products Division,Wescor Inc, Logan, UT

hFisher brand Superfrost plus, Fisher Scientific, Pittsburgh, PA

iAbcam, mouse monoclonal (SAP.4G5) to β1 tubulin, catalog number ab11312

j5% normal goat serum, 2.5% BSA in PBS, pH 7.4, Molecular Probes, Invitrogen, Carlsbad, CA

kMolecular Probes, goat anti-mouse IgG, 2 mg/ml, Alexa Fluor 488, catalog no. A11001, Molecular Probes, Invitrogen

lVectashield, 10 mL, catalog number H-1000, Vector Labs, Burlingame, CA

mNikon TE2000-E; camera Retiga EX; software Q-Capture Pro, Q Imaging, Vancouver, BC

n8.9 mL distilled water, 0.5 mL White's A, 0.5 mL White's B, 0.1 mL 10% glutaraldehyde, Electron Microscopy Sciences, Hatfield, PA

o6 mL distilled water, 0.5 mL White's A, 0.5 mL White's B, 3 mL 10% glutaraldehyde, Electron Microscopy Sciences

pLadd Research Industries Inc, Burlington, VT

qFEI Co, Hillsboro, OR; operated at 60 kV

rSAS release version 9.1, Cary, NC

sEpiInfo, Centers for Disease Control and Prevention, Atlanta, GA


The authors thank Dr Jim Wright, Dr Beth Spangler, Dr Sarah Fairchild, Dr John Dennis, and Mr Kevin King for assistance with statistical analyses, sample collection, and technical aspects of the study. The authors also thank owners and breeders of CKCS, including Mary Lynn and Ed Saidla and Charlsie Coe.

Brooke Davis was funded by the Morris Animal Foundation Veterinary Student Scholar Program for 2007.