Multiple-instance-learning-based detection of coeliac disease in histological whole-slide images.

We present a multiple-instance-learning-based scheme for detecting coeliac disease, an autoimmune disorder affecting the intestine, in histological whole-slide images (WSIs) of duodenal biopsies. We train our model to detect 2 distinct classes, normal tissue and coeliac disease, on the patch-level, and in turn leverage slide-level classifications. Using 5-fold cross-validation in a training set of 1841 (1163 normal; 680 coeliac disease) WSIs, our model classifies slides as normal with accuracy (96.7±0.6)%, precision (98.0±1.7)%, and recall (96.8±2.5)%, and as coeliac disease with accuracy (96.7±0.5)%, precision (94.9±3.7)%, and recall (96.5±2.9)% where the error bars are the cross-validation standard deviation. We apply our model to 2 test sets: one containing 191 WSIs (126 normal; 65 coeliac) from the same sources as the training data, and another from a completely independent source, containing 34 WSIs (17 normal; 17 coeliac), obtained with a scanner model not represented in the training data. Using the same-source test data, our model classifies slides as normal with accuracy 96.5%, precision 98.4% and recall 96.1%, and positive for coeliac disease with accuracy 96.5%, precision 93.5%, and recall 97.3%. Using the different-source test data the model classifies slides as normal with accuracy 94.1% (32/34), precision 89.5%, and recall 100%, and as positive for coeliac disease with accuracy 94.1%, precision 100%, and recall 88.2%. We discuss generalising our approach to screen for a range of pathologies.

Septic sialoadenitis in equids: a retrospective study of 18 cases (1998-2010).

REASON FOR PERFORMING STUDY: Septic sialoadenitis, although uncommonly reported in equids, is a significant cause of pain, inappetence, dysphagia and discomfort. There are currently few reported cases possibly as a result of its infrequent occurrence. OBJECTIVES: To review cases presenting with sialoadenitis and describe the presenting complaints, results of diagnostic tests, treatment and outcome. STUDY DESIGN: Retrospective case series. METHODS: Records were reviewed for equids presenting to the UC Davis William R. Pritchard Veterinary Medical Teaching Hospital between 1998 and 2010 for salivary gland swelling. Equids were included if a diagnosis of septic sialoadenitis was made based on a combination of oral examination and/or ultrasonographic findings and/or microbial culture. Data collected included age, breed, presenting complaints, diagnostic results, treatment and outcome. RESULTS: Eighteen equids were diagnosed with septic sialoadenitis affecting the parotid gland (11) or the mandibular salivary gland (7). Ultrasound was useful to differentiate whether the mandibular or parotid salivary gland was involved. Affected equids ranged in age from 4 to 30 years (mean 17.7 years). Fourteen of 15 (93.3%) equids that underwent a complete oral examination had dental or other oral abnormalities. Six of 18 cases had evidence of sialolithiasis. Culture of the infected salivary gland or secretions was performed in 9 equids and all yielded growth of Fusobacterium sp. along with other aerobic and anaerobic bacteria. Infection resolved in 15/18 cases (83.3%) and 2/18 (11.1%) were subjected to euthanasia. CONCLUSIONS: Dental disease and sialolith formation may play important roles in the development of septic sialoadenitis in equids. Anaerobic infection should be assumed in all cases and affected horses should be treated for this until culture and sensitivity results are available. Prognosis is favourable (83.3%) with appropriate treatment.

Characterization of Brucella abortus infection of bovine monocyte-derived dendritic cells.

Brucella abortus is a Gram negative facultative intracellular pathogen of cattle, and an important zoonosis in humans worldwide. Previous studies have shown that dendritic cells (DC) from humans and mice are highly permissive for Brucella survival and proliferation. Impairment of DC activation and maturation by Brucella infection has also been reported in these two species. The aim of this study was to characterize infection of bovine DC with B. abortus. Monocyte-derived DC (mdDC) were cultured from bovine peripheral blood mononuclear cells (PBMC) using the recombinant bovine cytokines IL-4 and GM-CSF. The resulting mdDC were DEC205(+), MHC class II(hi). Approximately 70% of the cultured cells were DEC205(+), MHC II(+). MdDC were infected with B. abortus strain 2308 at an MOI of 1 and 100. Parallel infection experiments were performed in monocyte derived macrophages (mdM) isolated from the same subjects. Bacteria were successfully killed by mdDC by 24 hours post infection (PI) with high and low dose of B. abortus, bacteria persisted in mdM infected with a high dose. Expression of IL-1b, IL-6, IL-10, IL-12p40, IFNγ, iNOS and TNFα in B. abortus infected and LPS stimulated mdDC at 6 and 24 hours PI were evaluated using RT-qPCR. At 6 hours PI all transcripts were increased over control cells and significantly less IL-10, IL-12p40, and IFNγ were expressed in mdDC infected with B. abortus compared to LPS stimulation. Evaluation of mdDC cultures by flow cytometry was performed. Flow cytometric analysis of infected and LPS stimulated mdDC 24 hours PI showed expression of CD80 and CD86 was impaired in two of the three animals analyzed. MHC class II expression was equivocal between the groups. From these results we conclude that cultured bovine mdDC are not permissive for intracellular proliferation of B. abortus, and infected mdDC exhibit some signs of maturational and activational impairment.

A potential role for indoleamine 2,3-dioxygenase (IDO) in Rhodococcus equi infection.

Rhodococcus equi is a facultative intracellular bacterial pathogen of foals and immunocompromised humans that infects and proliferates within host macrophages and dendritic cells (DC). Indoleamine 2,3-dioxygenase (IDO), the initial enzyme in the tryptophan catabolism pathway, is upregulated in R. equi infected equine monocyte-derived DC and alveolar macrophages. Tryptophan requirement of R. equi for extracellular and intracellular growth was assessed. Growth of R. equi in minimal media did not require tryptophan and pharmacologic inhibition of IDO had no effect on intracellular proliferation of R. equi in equine alveolar macrophages. To investigate an immune-regulatory role for INDO in R. equi infection, IDO(-/-) (B6.129-(Indotm1Alm)/J) (n=22) and strain matched control (C57BL/6J) (n=20) mice were infected with R. equi by intraperitoneal injection, for 3 and 6 days. There was no difference in bacterial counts in liver or spleen between the two groups. Histological sections of liver and spleen were assigned inflammation scores and RT-PCR for interferon-gamma (IFNγ), tumor necrosis factor-alpha (TNFα), IL-4, IL-6, IL-10, IL-12, IL-23, forkhead box P3 (FoxP3), and transforming growth factor-beta (TGFß) was performed on liver and spleen. Liver tissue of IDO(-/-) had higher inflammation scores at 6 days post-infection (PI) (P=0.05) and had decreased expression of TGFß at 3 days PI (P=0.01), and FOXP3 at 3 days (P=0.02) and 6 days (P=0.03) compared to control mice. Immunostaining for FOXP3 showed lower numbers of FOXP3+ regulatory T cells in liver of IDO(-/-) mice 6 days PI. Prolonged inflammation in the liver tissue of IDO(-/-) mice corresponded with lower expression of FOXP3 and TGFß in that tissue, and also with lower numbers of FOXP3+ regulatory T cells. We conclude that IDO expression by activated macrophages and DC plays a role in dampening the inflammatory response to R. equi infection in mice.

Identification of immunologically relevant genes in mare and foal dendritic cells responding to infection by Rhodococcus equi.

Rhodococcus equi is a facultative intracellular bacterial pathogen of horses; infected foals develop pyogranulomatous pneumonia, however adult horses are largely unaffected. R. equi infects and proliferates within host macrophages and dendritic cells (DCs). DCs initiate the appropriate adaptive immune response, thereby playing a critical role in determining the outcome of infection. Our aim was to identify genes that are differentially expressed in R. equi infected monocyte-derived DCs (mdDCs). Peripheral blood monocytes from mares and foals were used to derive mdDCs by culturing with recombinant equine IL-4 and recombinant human GM-CSF. RNA harvested 24h after infection with R. equi (ATCC 33701+) was used to perform suppression subtractive hybridization (SSH) experiments. Approximately 38 unique sequences were obtained from these experiments. Differential expression of 19 immunologically relevant genes was validated by PCR. These genes are characterized by the following functions: cell adhesion, chemotaxis/migration, immune/inflammatory response, ion transport, signal transduction, T-cell regulation, and vesicular transport. In summary, we identified several novel genes that are differentially expressed in foal and adult mdDCs in response to R. equi infection. These genes provide promising targets for further research into the host response to R. equi, and the susceptibility of foals to this disease.

Myopathy in horses with pituitary pars intermedia dysfunction (Cushing's disease).

Fifteen horses with pituitary pars intermedia dysfunction were studied. The horses were of various breeds and between 15 and 28 years of age. Control horses matched for breed and age were studied for comparison. Evaluations included complete blood cell count and serum biochemical analysis, electromyography, and gluteus medius muscle biopsies for histochemical, morphometric, and ultrastructural analysis. No differences were found between groups of horses on routine laboratory analysis or electromyography. We demonstrated that muscle wasting in diseased horses was the result of atrophy of types 2A and 2B muscle fibers and loss of type 2B myofibers. Mild non-specific non-inflammatory myopathic alterations such as myofiber size variation, internal nuclei, perimysial, endomysial and sarcoplasmic fat accumulation were observed. At the ultrastructural level, subsarcolemmal mitochondrial accumulation and increased lipid droplets were evident. Similar to other species, this study confirmed atrophy of type 2 fibers as the cause of muscle mass loss in horses with Cushing's disease.

Genomic characterization of equine interleukin-4 receptor alpha-chain (IL4R).

Three overlapping fragments of the equine interleukin-4 receptor alpha chain gene (IL4R) were cloned and sequenced. The resulting 3553 bp cDNA sequence exhibited homology to human, murine and bovine IL4R. The equine IL4R exhibits many conserved features when compared to other species, including intron-exon boundary positions and amino acid sequence motifs characteristic of type I cytokine receptors. The IL4R gene was localized to horse chromosome ECA13 by synteny mapping on a somatic cell hybrid panel. Evidence for an alternative splice variant of IL4R was found in the genomic sequence and subsequently verified using RT-PCR on equine monocyte RNA. A polymorphism screen of the largest exon, homologous to exon 12 of the human IL4R gene, was performed using DNA from 60 horses of various breeds which yielded 11 coding-region single nucleotide polymorphisms (SNPs), 7 synonymous and 4 non-synonymous. Three of the four non-synonymous SNPs occur at high frequencies and are found very near a conserved tyrosine residue.

Purpura haemorrhagica in 53 horses.

The medical records of 53 horses with purpura haemorrhagica were reviewed. Seventeen of them had been exposed to or infected with Streptococcus equi, nine had been infected with Corynebacterium pseudotuberculosis, five had been vaccinated with S. equi M protein, five had had a respiratory infection of unknown aetiology, and two had open wounds; the other 15 cases had no history of recent viral or bacterial infection. The horses were between six months and 19 years of age (mean 8.4 years). The predominant clinical signs were well demarcated subcutaneous oedema of all four limbs and haemorrhages on the visible mucous membranes; other signs included depression, anorexia, fever, tachycardia, tachypnoea, reluctance to move, drainage from lymph nodes, exudation of serum from the skin, colic, epistaxis and weight loss. Haematological and biochemical abnormalities commonly detected were anaemia, neutrophilia, hyperproteinaemia, hyperfibrinogenaemia, hyperglobulinaemia and high activities of muscle enzymes. All of the horses were treated with corticosteroids; 42 also received non-steroidal anti-inflammatory drugs and 26 received antimicrobial drugs. Selected cases received special nursing care, including hydrotherapy and bandaging of the limbs. Most of the horses were treated for more than seven days and none of them relapsed. Forty-nine of the horses survived, one died and three were euthanased, either because their severe clinical disease failed to respond to treatment or because they developed secondary complications. Two of the four non-survivors had been vaccinated against S. equi with a product containing the M protein, one had a S. equi infection and the other had a respiratory infection of undetermined aetiology.

Evaluation of the SNAP foal IgG test for the semiquantitative measurement of immunoglobulin G in foals.

The SNAP Foal IgG test (IDEXX) as evaluated for its accuracy and usefulness by measuring blood samples collected from 42 foals between 24 and 48 hours after birth. The results were compared with the single radial immunodiffusion (SRID) test as the reference method. The SNAP test was quick and easy to perform, and the results were similar to those obtained by SRID in 64 per cent of the samples. The best results were found with low (< 400 mg/dl) and high (> 800 mg/dl) concentrations of immunoglobulin G, with an accuracy of 80 per cent and 89 per cent, respectively. The intermediate concentrations were usually lower when measured by the SNAP test than by the SRID test, possibly owing to the variable volume of blood added to the test with the sample loop.

Marsupialization and iodine sclerotherapy of a branchial cyst in a horse.

A 6-month-old Morgan colt was evaluated because of a 10-cm right-sided retropharyngeal swelling. The swelling was soft and moveable on examination, and palpation did not elicit signs of pain. Radiography revealed a large space-occupying mass ventral to the second cervical vertebra; ultrasonography revealed an anechoic fluid-filled structure with a well-defined hyperechoic capsule. Fine-needle aspiration yielded a viscous amber fluid. Cytologic evaluation indicated that the fluid was an exudate; anaerobic and aerobic bacterial culture did not yield any growth. Histologic examination of a portion of the cyst capsule revealed a connective tissue wall lined by pseudostratified columnar to cuboidal epithelium, consistent with a branchial cyst. The cyst wall was marsupialized to the skin, and iodine sclerotherapy was performed twice daily for 14 days, at which time forceps were introduced into the cyst and the cyst lining was removed. The site was allowed to heal by second intention, but 10 days later, the swelling recurred. An incision was made over the previous marsupialization site, and residual remnants of the cauterized cyst lining were removed with a forceps. The foal did not have any other complications during the subsequent 2 years. Branchial arch cysts are uncommon embryonic anomalies of horses, mice, cats, dogs, and cattle. Results suggest that marsupialization and iodine sclerotherapy may be a viable alternative to surgical excision in horses with branchial cysts; however, the entire cyst lining must be removed at the completion of sclerotherapy to prevent recurrence and abscess formation.

Molecular cloning and sequencing of the low-affinity IgE receptor (CD23) for horse and cattle.

Expression of the low-affinity IgE receptor (CD23) on the surface of mononuclear cells is a critical event in the development of IgE-mediated immunologic responses. Using PCR and cDNA library screening, a 2.7kb cDNA encoding equine CD23 and a 627bp PCR fragment of cattle CD23 were sequenced and characterized. The equine CD23 sequence encodes a complete and continuous open reading frame of 327 amino acids, while the shorter cattle fragment encodes 209 amino acids corresponding to nucleotides 325-1098 of the equine CD23 transcript. In addition to high identities in their nucleotides and translated amino acids with CD23 sequences published for other species, the translated equine CD23 protein also shares many of the structural features of this molecule described for human and rodents. Interestingly, an additional repetitive element of possible functional significance consisting of 18 amino acids, located between the transmembrane region and the carbohydrate-binding lectin domain of horse CD23, was also identified. Based on these sequences, molecular assays will be developed to measure CD23 mRNA in tissues and expression of recombinant proteins will be essential to the production of species-specific antibody reagents. These assays and reagents will be useful in future studies of allergic and lymphoproliferative diseases in horses and cattle.

Injection site eosinophilic granulomas and collagenolysis in 3 horses.

Three horses were presented with a history of having developed raised cutaneous nodules, within 24-48 hours, in areas of previous injections using standard silicone-coated hypodermic needles. Skin biopsies were taken from a selected cutaneous nodule from all horses for histopathologic evaluation. Histologically, the nodules were consistent with a diagnosis of equine eosinophilic granuloma. A hypersensitivity reaction to the silicone, or another component of the coating formulation, was hypothesized to be responsible for these lesions. Two horses were experimentally injected using both coated and noncoated stainless steel hypodermic needles and skin biopsies were obtained 14 days after injection. The sites of the coated needle injections were characterized by severe eosinophilic granulomatous inflammation with and without collagenolysis. The eosinophilic granulomas with and without collagenolysis observed in these horses are proposed to represent a complex immunologic response to the silicone-based coating of most hypodermic needles.

Phenotypic characterization of lymphocyte subpopulations in horses affected with chronic obstructive pulmonary disease and in normal controls.

The alterations in lymphocyte subsets in chronic obstructive pulmonary disease (COPD) in the horse were investigated by using monoclonal antibodies to identify CD3+, CD4+, CD8+, and surface immunoglobulin positive (sIg+) lymphocytes in peripheral blood, bronchoalveolar lavage fluid (BALF), and pulmonary biopsy frozen tissue sections. COPD-affected horses (n = 5) and normal controls (n = 5) were sampled prestabling and 14 days poststabling, at which time the COPD-affected horses wee exhibiting clinical signs of COPD. The peripheral blood absolute CD4+ lymphocyte count was significantly elevated in the COPD-affected horses pre- and poststabling. The CD4:CD8 ratio in peripheral blood of COPD-affect horses was unaffected by stabling, but the same ratio in the control horses was significantly decreased. These findings support a hypothesis of deficient regulation of a systemic immune response to indoor air in the COPD-affected horses. A large population of leukocytes in pulmonary biopsy immunohistochemical sections from both groups of horses appeared to be CD3+ CD4- CD8-, an uncommon phenotype in both the peripheral blood and BALF.

Effect of treatment with erythromycin on bronchoalveolar lavage fluid cell populations in foals.

OBJECTIVE: To determine whether oral administration of erythromycin alters the inflammatory response to bronchoalveolar lavage (BAL) in young horses. ANIMALS: 12 healthy, unweaned, mixed-breed foals of either sex, between 2 and 4 months old. PROCEDURE: BAL was performed; 250 ml of phosphate-buffered saline solution (300 mOsm, pH 7.4) was administered in 50-ml aliquots. Foals were carefully monitored for 4 days, then erythromycin base (25 mg/kg of body weight, PO, q 12 h) was given to foals of the treated group. After 4 days, foals were reanesthetized, and the same lung was relavaged. Cytologic examination was performed on BAL fluid (BALF) samples from both groups of foals. At 12 hours after administration of the final dose, erythromycin A and anhydroerythromycin A concentrations were determined in plasma of treated foals. RESULTS: In the second BALF sample from the same lung of control foals, percentage of neutrophils was significantly increased (53 +/- 38.0%) [corrected], compared with that from erythromycin-treated foals (4.88 +/- 3.66%, P < 0.05), and was associated with apparent decrease in the ability of BALF cells from erythromycin-treated foals to migrate toward a chemoattractant source. Significantly fewer BALF cells adhered to a cell culture substratum after erythromycin treatment of foals. Erythromycin A was not detected in plasma of any treated foal at the time of the second BAL; anhydroerythromycin A, a degradation product of erythromycin, was detected in plasma of 5 of 6 foals (mean concentration, 0.2 +/- 0.06 micrograms/ml). CONCLUSION AND CLINICAL RELEVANCE: Bal induces neutrophilic inflammation, which persists for at least 4 days in the lungs of young horses. Erythromycin [corrected] (25 mg/kg, PO q 12 h) diminishes this inflammatory response through a mechanism that may involve alteration of BALF cell function. Degradation of erythromycin to biologically active products or presence of parent drug in pulmonary secretions may be responsible for alterations in pulmonary lavage cell chemotaxis and adherence. Erythromycin administered orally to foals at clinically relevant doses appears to have nonantimicrobial effects that may interfere with host cell metabolism and decrease inflammatory responses in airways.

Characterization of horse (Equus caballus) T-cell receptor beta chain genes.

Genes encoding the horse (Equus caballus) T-cell receptor beta chain (TCRB) were cloned and characterized. Of 33 cDNA clones isolated from the mesenteric lymph node, 30 had functionally rearranged gene segments, and three contained germline sequences. Sixteen unique variable segments (TCRBV), 14 joining genes (TCRBJ), and two constant region genes (TCRBC) were identified. Horse TCRBV were grouped into nine families based on similarity to human sequences. TCRBV2 and TCRBV12 were the most commonly represented horse families. Analysis of predicted protein structure revealed the presence of conserved regions similar to those seen in TCRB of other species. A decanucleotide promoter sequence homologous to those found in humans and mice was located in the 5' untranslated region of one horse gene. Germline sequences included the 5' region of the TCRBD2 gene with flanking heptamer/nonamer recombination signals and portions of the TCRBJ2-C2 intron. Southern blot hybridizations demonstrated restriction fragment length polymorphisms at the TCRBC locus among different horse breeds.

Variables related to body-weight status of mentally retarded adults.

Mentally retarded male and female adult subjects displayed mean body weights in excess of their ideal weights; excessive body weight of females was also apparent in comparison to normative data for the United States. Maintenance of appropriate weight appeared to be more likely in a controlled residential setting than in the natural home environment. These sex and environmental relationships could not be explained by medication and dietary programming differences, and age, race, and level of retardation were unrelated to body weight.

Pulmonary candidiasis complicating a severe burn treated with miconazole.

A 33/4-year-old boy received a 50 per cent body surface area burn in a house fire. On the eighteenth day post burn he was moribund and a diagnosis of pulmonary candidiasis was made following tracheal cultures and X-ray changes after a period of progressive deterioration during which he had received multiple antibiotics and assisted ventilation. Treatment of this pulmonary candidiasis was with intravenous miconazole. The child's clinical condition rapidly improved over the following 24-48 hours and the X-ray changes resolved. He was subsequently discharged fully healed.

Evidence for the bilobal nature of diferric rabbit plasma transferrin.

Plasma transferrin is involved in iron transport within the circulatory system of vertebrates, and provides an iron source for haemoglobin synthesis and other metabolic requirements. However, despite extensive studies by spectroscopic, biochemical and physiological techniques, the nature of iron binding and the mechanisms of uptake and release of iron are not fully understood. Plasma transferrins are monomeric glycoproteins with a molecular weight of approximately 80,000 (ref. 2); they have two similar and very strong binding sites for Fe(III), together with two associated anion binding sites. Fragmentation studies on various transferrins have shown that the polypeptide chain is composed of two domains formed from the N-terminal and C-terminal halves of the polypeptide chain. Each domain contains one metal binding site. The marked sequence similarities which exist between the two halves may reflect a doubling of an ancestral structural gene during the phylogenetic development of the protein. Preliminary crystallographic investigations of diferric rabbit plasma transferrin have been reported from this laboratory. We now report initial studies of the X-ray structure determination of dife-ric rabbit plasma transferrin which have led to a 6-A resolution electron density map.