Case Report Commensal and multidrug-resistant Neisseria spp. sepsis in feline

Introduction: Sepsis is a serious problem in felines with a mortality rate ranging from 29-79%. Neisseria spp. is considered a commensal microorganism of the oral cavity of dogs and cats and is usually isolated from human wounds resulting from bites of these animals. Case Report: The present report describes clinical, imaging and laboratory findings of a feline with sepsis wherein commensal and multidrug-resistant (MDR) Neisseria spp. was isolated. The feline presented a history of four days of anorexia, dyspnea, prostration, and, pericardial, pleural and abdominal effusions. Pericardiocentesis was performed and hemorrhagic exudate was observed. The animal died after 11 days of treatment with gentamicin and amoxicillin combined with clavulanic acid. During necropsy, the abdominal cavity was found to be filled with greenish-yellow content and the pericardial sac was thickened with a large amount of purulent secretion. Histopathology revealed sepsis with necrotizing suppurative pericarditis, diffuse mononuclear pneumonia and necrotic pleuritis, leading to secondary bacterial infection. Conclusions: Commensal Neisseria spp. are important zoonotic bacteria, which trigger a serious disease in felines. However, it has not been reported to cause sepsis with pneumonia, suppurative necrotizing pericarditis and pericardial effusion.

Abdominal ultrasound showed hepatosplenomegaly, loss of renal corticomedullary definition, and a small amount of free fluid. Radiographic assessment of chest showed a dorsally displaced trachea with preserved luminal diameter and a marked increase in cardiac silhouette topography that has a globular appearance. Pulmonary fields presented an alveolar pattern with interlobar fissures in the right caudal lobe. Diaphragmatic line was intact although partially obliterated ( Figure 1). Such images suggested pericardial and pleural effusion. The pericardium was thicker and had moderate pericardial effusion and observed through the Doppler echocardiograph.
Cytological and biochemical analysis of peritoneal, pleural, and pericardial effusions were performed according to Stockham et al. [10]. Pericardial effusion was classified as hemorrhagic exudate, and had a cloudy and whitish appearance with positive coagulation, pH 7; glucose was absent, occult blood was present, and a total of 240.4 × 10 3 µL nucleated cells were observed. It was not possible to perform density and protein analysis due to the condition of the sample. Cytology showed leukocytes prevailed, with 86% neutrophils, 8% lymphocytes, 6% macrophages and abundant red blood cells Normal and reactive mesothelial cells (3-6/field) and countless degenerated cells were observed.
Additionally, the pericardial effusion culture was processed in the following media: 8% sheep blood agar (Sigma-Aldrich, Darmstadt, Germany), MacConkey agar (Neogen Corporation, São Paulo, Brazil) and Sabouraud Dextrose agar (Sigma-Aldrich, Darmstadt, Germany) under aerobic conditions, incubated at 37 °C for 72 hours. The colonies were identified according to Markey et al. [11], through morphology, Gram stain, catalase test, oxidase test, and biochemical series (Triple Sugar Iron-TSI, Sulphide Indole Motility -SIM, OF-GOF Test, Citrate, Urea, and Gelatin) (Sigma-Aldrich, Darmstadt, Germany). The isolate had a nonhemolytic and non-lactose fermenter yellow colony, identified as a Gram-negative diplococcus, and catalase and oxidase positive. The biochemicals series were inert, gelatin and urea were negative.
A sample colony was inoculated in Brain-Heart infusion broth (BHI -Himedia Labs. Mumbai, India) and incubated at 37 °C overnight. Genomic DNA was extracted by the phenol-chloroform method according to Sambrook and Russel [12].
The extracted DNA was used to amplify the 16S ribosomal RNA (rRNA) through polymerase chain reaction (PCR). The oligonucleotide pair sequences used were 27F: AGA GTT TGA TCC TGG CTC AG [13] and 1492R: GGT TAC CTT GTT ACG ACT T [14] which amplify a 1512 base pair fragment.
The reactions were amplified in a MyCycler TM thermal cycler (Biorad, California, USA) with initial denaturation of 5 minutes at 95 °C, followed by 35 denaturation cycles for 45 seconds at 95 °C, hybridization for 1 minute at 52 °C and 1 minute and 30 seconds at 72 °C extension, and a final extension cycle at 72 °C for 7 minutes. The PCR products were separated on 1.0% agarose gel, stained with GelRed TM (Biotium®, UK) at 10 V/cm and visualized on a photodocumenter. Subsequently, the product obtained by PCR was purified and sequenced in the ABI 3500 Genetic Analyzer automatic sequencer (Applied Biosystems Foster City, CA, USA), according to the manufacturer's recommendations. The sequence was deposited with access number MN998616 and compared in the GenBank database, using BLAST on the NCBI server (http:www.ncbi.nlm.nih.gov/BLAST) and a threshold value of > 98.6% similarity to 16S rRNA of a given species was used to differentiate two species [15].
We processed the sequence using CLC DNA Workbench Program (6.0) and analyzed with Basic Local Alignment Search Tool (BLAST) to verify the similarity with other corresponding sequences available on GenBank. The analysis was performed on the Phylogeny.fr platform and the sequence generated in PCR was aligned using the MUSCLE program The phylogenetic tree of the isolate was generated by the Maximum Likelihood method using the PhyML software (v3.1/3.0 aLRT) with the HKY85 substitute model, selecting an estimated proportion of invariant locations (from 0.000) and 4 rate categories distributed by range to explain the heterogeneity of rate between sites. The Gamma parameter was estimated directly from the data (gamma = 1.111). The aLRT (SH-Like) test was used for assessing internal branch reliability. Phylogenetic tree was edited and graphically represented using the TreeDyn (v198.3). The homologous sequence of Moraxella canis (NR028914.1) HSP gene was included in phylogenetic tree as an external group (outgroup) (Figure 2).
After confirmation of pleural and pericardial effusions, thoracentesis was performed. This was followed by pericardiocentesis. Additionally, amoxicillin with clavulanic acid (20 mg/kg/IV every 12 hours) antimicrobial therapy, ondansetron hydrochloride (0.22 mg/kg/IV every 8 hours), omeprazole (1 mg/kg/IV every 24 hours), meloxicam (0.1 mg/kg/IV every 24 hours), multivitamin supplement (4 ml/IV every 24 hours) and lactated Ringer's solution were prescribed. Gentamicin was added (4 mg/kg/IV every 12 hours) to the treatment based on the antimicrobial susceptibility test. On the second day, the feline accepted pasty food. On the fourth day, clinical signs did not improve, and radiography revealed the recurrence of pleural effusion; a new thoracentesis was performed. On the seventh day, the animal developed hypothermia ± 36.3 °C and dyspnea, and despite all the measures taken, the feline died on day 11.
During necropsy, we observed severe periodontal disease with formation of bacterial plaques in the canines, premolars, and molars, in addition to tooth mobility. There were approximately 50 mL of translucent brownish liquid with filaments of proteinaceous material in the abdominal cavity. The liver had a moderately lobular pattern. We also found a Pericardial sac with marked dilation and yellowish color. Both organs with deposition of material with firm yellowish and fibrillar consistency, forming easily detachable plaques from the surface. B. Macroscopic analysis of opening the feline pericardial sac at necropsy. Thickened pericardial sac with a large amount of purulent-looking secretion that strongly adheres to the epicardium. C. Feline lung microscopic section shows pleuritis associated with Neisseria spp. Moderate pleural thickening with mixed infiltrate and deposition of amorphous hyaline material on the surface. The hematoxylin-eosin stain was used. D. Feline heart microscopic section associated pericarditis with Neisseria spp. Pericardium is markedly thickened by accentuated deposition of amorphous material, caused of cellular debris and filamentous material (fibrin), and marked by the mixed inflammatory infiltrate composed of degenerate and healthy neutrophils, and foamy macrophages. moderate amount of yellow-brown liquid in the thoracic cavity. The lung was hepatized and diffusely reduced, and the left lobes had adhered to the thoracic pleura. Lung and pericardial sac diffusely showed deposition of material with a firm yellowish and fibrillar consistency, formed plaques easily that were detachable from the pleural surface ( Figure 3A). Upon inspection of the pericardial sac, an accented dilation with a pasty content yellowish (purulent aspect) was noted that strongly adhered throughout the epicardial layer ( Figure  3B). The walls of the ventricles were thicker and other organs had no macroscopic changes.
The pericardium was observed to be moderately thick upon histological evaluation and covered by an accentuated amount of amorphous and filamentous material (fibrin), cellular debris, and marked mixed inflammatory infiltrate, composed of degenerate and healthy neutrophils, foamy macrophages, and occasional giant cells. Furthermore, there were multifocal areas of random distribution with aggregates of amorphous material, strongly basophilic, and areas of calcification ( Figure 3D). Alveolar septa were diffusely thickened and infiltrated by an accentuated amount of mononuclear inflammatory cells and moderately diffuse congestion. The visceral pleura was moderately thickened with multifocal coalescent areas of necrosis ( Figure 3C). Other multifocal areas of fibrin deposition, necrotic material, cellular debris, and a mixed inflammatory infiltrate were similarly found in the epicardium. We observed in the meninge a deposition of fibrillar eosinophilic material, occasionally adhered to the vessel wall, interspersed with red blood cells, and rare inflammatory cells (thrombi).
According to these findings, sepsis with pericarditis and diffuse purulent fibrin pleuritis was confirmed, secondary to bacterial infection by Neisseria spp.

Discussion
The genus Neisseria cannot be differentiated based on phenotypic characteristics and basic biochemical tests. Thus, to complete a more reliable phylogenetic analysis the molecular tests must be carried out using different genes. Interspecies diversity varies from 1 to 5% for 16s rRNA gene, which is insufficient for a phylogenetic resolution. Therefore, other genes must be included, such as rpoB, for genetic identification [4,19]. The 16s rRNA gene product amplified from the pericardium shared greater sequence similarity with N. animolaris-MH166779.1 (99.63%) and N. canis-AY426974.1 (99.63%), making it impossible to distinguish the exact species.
The main habitat of Neisseria spp. is the oral cavity, and can be found in man and domestic animals. In dogs, these bacteria are present in the saliva, nasopharynx, throat, dental plaque and nasal fluids [20,21], and in cats, both sick and healthy, the bacteria are present in the oral cavity, since it is a commensal microorganism of the oral cavity [22]. In the current study, the feline had a severe periodontal disease and this is considered the probable gateway of Neisseria spp to hematogenous pathway.
Cases of infection in humans by Neisseria spp. are associated with dog and/or cat bites that carries the bacteria. However, Allison and Clarridge [23] described an isolated case of acquired infection of N. canis by aerosols from a carrier dog in a human patient with chronic bronchiectasis, who was diagnosed through bacterial isolation in sputum. The first report about N. canis was published in 1962 where the researchers isolated the bacterium from the pharynx of a healthy dog [24]. After, the first isolation of N. canis in a dog mandibular abscess was described, future hypotheses were based on the origin of the N. canis in an abscess which facilitates penetration of a foreign body in to the oral mucosa [4].
Neisseria animaloris can also cause systemic infections in humans and animals [25] and recently Foster et al. [26] reported recovery of N. animaloris from stranded harbor porpoises (Phocoena phocoena) in northern Europe. The referral study showed that these infections occurred after traumatic injury caused by predators (gray seals -Halichoerus grypus). The ultimate cause for death of the porpoises was widespread bacterial infection. There were several lung abscesses with caseous exudates, suggesting hematogenesis. Despite the absence of a history of biting by another animal, the reported feline case refers to the spread of bacterial infection, triggering sepsis.
The main morphological findings were pericarditis and fibrinopurulent pleuritis related to clinical changes and the exudate accumulation in pericardial sac with a cardiac silhouette increase in this feline. The mechanisms that are responsible for the occurrence of pericarditis include hematogenous dissemination, direct spread of an intrathoracic infection such as pneumonia, trauma, and thoracic surgery [27]. In this case, due to the concomitant pericarditis, pleuritis and peritonitis, hematogenous dissemination leading to sepsis is the most likely cause, even though primary infection cannot be identified, as in most veterinary reports [3].
Lobetti [28] reported a case of a feline with bacterial pericarditis caused by Peptostreptococcus which was determined after dental prophylaxis. On the other hand, Pasteurella multocida, Actinomyces canis, Fusobacterium, and Bacteroides were isolated in pericardial effusion from a feline with unknown etiology, and was successfully treated with antibiotic therapy [3]. A similar case was reported resulting from infection by Moraxella osloensis and Bacteroides, both considered commensals of feline oral cavity [29]. Nevertheless, the animal was unresponsive to treatment and died after seven days. There are reports of Neisseria meningitidis in humans, causing bacterial meningitis, infecting extra meningeal sites such as the pericardium, and leading to cardiac tamponade [30].
The feline in the present report had severe periodontal disease, pneumonia, pleuritis and pericarditis. It is likely that the hematogenous spread of Neisseria spp. caused sepsis, since the animal had no history of bites, trauma or thoracic surgery. A limitation of this report is that the feline has not been tested for FIV (Feline Immunodeficiency Virus) and FeLV (Feline Leukemia Virus) that are known to cause feline immunosuppression. There was no morphological evidence of infection by feline Coronavirus which has been found to cause serous effusions.
Sepsis therapy involves antibiotics and hemodynamic support, among other therapies [31], but veterinary literature, especially in cats, is still scarce. In cases of bacterial pericarditis, therapy consists of pericardiocentesis to correct cardiac tamponade and antibiotic therapy [26] through the susceptibility test.

Conclusions
Neisseria spp. can cause serious diseases in felines and have never been previously reported in the literature. Neisseria spp. causes sepsis with pneumonia, suppurative necrotizing pericarditis and pericardial effusion. N. canis and N. animaloris are bacteria of zoonotic importance present in the oral cavity of felines and can offer risks of transmission and infection to humans.