Courtesy of: www.ivis.org

 

In:  Recent Advances in Canine Infectious Diseases, Carmichael L. (Ed.). International Veterinary Information Service, Ithaca NY (www.ivis.org), Last updated: 29-Apr-2005; A0105.0405

Canine Coronavirus Infection

A. Pratelli

Dept of Animal Health and Well-being, Univeristy of Bari, Valenzano, Bari, Italy.

Introduction

Canine coronavirus (CCoV) is a cause of sporadic outbreaks of enteritis in dogs. The infection was first described by Binn in 1971 during an epizootic in Germany, although serological evidence was obtained earlier suggesting that dogs may be naturally infected with a coronavirus related to swine transmissible gastroenteritis virus (TGEV). CCoV is now recognized as an important, but imperfectly understood, pathogen of dogs. CCoV appears to be enzootic worldwide since the virus has been demonstrated in Europe, United states, Thailand and Australia. Studies performed on sera collected from both normal dogs and dogs in outbreaks of enteritis in kennels in the Netherlands, revealed a high rate of antibodies to CCoV in both categories. Seroprevalence studies in several countries indicate that prevalence rates vary from 0 to 80%; in one large study, 45% seroprevalence rates were reported in normal dogs, contrasting with 61% in diarrheic dogs. The seroprevalence appeared to depend on the population of dogs tested; generally higher rates were observed in endemically infected kennels, rather than in family dogs or random source laboratory dog.

Antigenic Groups

Coronaviruses can be divided into three distinct antigenic groups; within each group, the viruses are classified according to their natural hosts, nucleotide (nt) sequences and serologic relationships. There are two groups of mammalian coronaviruses, groups I and II, while the two avian coronaviruses, infectious bronchitis virus (IBV) and turkey coronavirus (TCoV), form the third group.
Group I includes the prototype, human coronavirus (HCoV) strain 229E, and the coronaviruses of pigs (transmissible gastroenteritis virus, TGEV, porcine respiratory coronavirus, PRCoV, and porcine epidemic diarrhea virus, PEDV), cats (feline coronaviruses, FCoVs) and dogs (canine coronaviruses, CCoVs). On the basis of their relationship to CCoV, FCoVs can be distinguished in two serotypes, FCoV type I and FCoV type II. By sequence analysis, it was demonstrated that the two prototypes of FCoV type II (strains 79-1146 and 79-1180) originated from double and separate recombination events between FCoV type I and CCoV. Since it was observed that the spike protein had been acquired from CCoV, it is understandable why FCoV type II viruses are serologically related to CCoV.
The second group of mammalian coronaviruses includes the prototype murine hepatitis virus (MHV) and the coronaviruses of humans (HCoV-OC43 and HECoV-4408), rats (RCoV), cattle (BCoV), pigs (PHEV) and other species. In 2002 a new coronavirus was discovered in association with outbreaks of a severe acute respiratory syndrome (SARS) in humans. The complete genomic sequences of several SARS-CoVs was determined, revealing that this coronavirus is not closely related to any of the previously characterized animal and human coronaviruses. Consequently, it was proposed to include this virus in a new antigenic group.

Etiology

Coronavirus, a genus in the family Coronaviridae, are large, enveloped, positive-stranded RNA viruses responsible for highly prevalent diseases in humans and domestic animals (Fig. 1). The RNA genome associates with the nucleocapsid (N) phosphoprotein to form a helical nucleocapsid. Negatively stained electron micrographs of coronavirus particles show the distinct corona effect produced by the pedunculated surface projections which are approximately 12 - 24 nm long and have a club-shaped end about 10 nm wide. The surface projections are spaced widely apart and dispersed evenly over all the surface. Like other enveloped viruses, CCoV is sensitive to heat, lipid solvents, non-ionic detergents, formaldehyde and oxidizing agents.
The virus is acid-stable and can be stored for years at -70°C or lyophilized at +4°C. Coronaviruses have the largest genomes, 27 to 32 kb, of all RNA viruses and replicate by a unique mechanism, which results in a high frequency of recombination. The genome includes 7 to 10 open reading frames (ORFs) encoding both structural and non-structural proteins. Gene 1 consists of two overlapping regions (ORF1a and ORF1b), and is translated into a polyprotein, a precursor of the viral replicase (Rep). Downstream from the Rep gene there are four structural proteins, 5’-Rep-S-E-M-N-3’, interspersed with several ORFs encoding various non-structural proteins and, in some strains, the HE glycoprotein, which differs markedly among coronaviruses in number, nt sequence, gene order and methods of expression.
The S glycoprotein (150 - 200 kd) forms the large petal shaped spikes on the surface of the virions and can be divided into three structural domains: a large external domain that is further divided into two sub-domains, S1 and S2, a transmembrane domain and a short carboxyl-terminal domain. The S1 sequence is variable and mutations in the S1 sequence have been associated with altered pathogenicity and antigenicity of the virus; in contrast, the S2 sequence is more conserved. The cleavage into S1 and S2 depends mainly on the host cell type and may enhance cell-fusion activity or viral infectivity; however, the uncleaved S protein may mediate cell fusion, though at a lower efficiency. The S protein of coronaviruses group I is not cleaved.
The S glycoprotein regulates several important biologic functions: attachment to cells, fusion of the viral envelope with host cell membranes and, sometimes, cell to cell fusion and is the major inductor of virus-neutralizing antibodies. The M glycoprotein (about 29 kd) differs from the other proteins because only a short amino-terminal domain of the protein is exposed out of the viral envelope. This domain is followed by a triple-membrane-spanning domain and a large carboxyl-terminal domain inside the envelope. Although the major immunological role has been attributed to the S protein, both the amino- and the carboxyl-termini of the M protein elicit strong immune response inducing antibody-dependent, complement-mediated, virus neutralization. The small envelope protein E, which is 9 to 12 kd, was recently shown to be associated with the viral envelope. Together with the M protein, E protein is required for budding of virions. The nucleocapsid protein N is a phosphoprotein of 50 to 60 kd that interacts with viral genomic RNA to form the viral nucleocapsid. The N protein in the viral nucleocapsid further interacts with M protein leading to the formation of virus particles.

Coronavirus01

Figure 1. Characteristic canine coronavirus (EM) particles from stool of dog with diarrhea.

CCoV Molecular Evolution

In the last few years, important data on the diffusion and the biology of this virus have been obtained. Several recent papers have demonstrated genetic divergence from the reference CCoV strains. This was evident by the presence of variant CCoV genomes identified in fecal samples of infected dogs. Sequence analysis of the M and S genes of several CCoV-positive fecal samples, demonstrated that a novel genotype (Elmo/02) circulates among dogs. This new virus reveals high prevalence of nt and amino acid (aa) identity with FCoV type I and, consequently it has been proposed to designate it as CCoV type I and the classical CCoV strains as CCoV type II.
The two genotypes of CCoV diverge substantially in the S gene, where there is greater than 39% nt and about 46% aa variation. Analysis of the S gene of different CCoVs strains (Elmo/02-like) revealed only small variation (4 - 8%), probably as consequence of their different geographical origin and most of the sequence changes observed are conservative, demonstrating that there is some heterogeneity in the S gene of these new viruses. The presence of a stretch of basic residues RRXRR suggests a possible cleavage of the protein. A similar basic motif is present in the same approximate position in all group II and III coronaviruses identified and classified to date, but is absent in Group I coronaviruses. Cleavage of the S protein of coronaviruses has been correlated to cell-fusion activity in-vitro, but potential implications in the pathobiology of the virus have not been determined.

Available data suggest that the two genotypes of CCoV have undergone a linear evolution rather than a sudden shift originating from a recombinant event analogous to those leading to the appearance of FCoVs type II. A possible explanation for this phenomenon is that homologous recombination events between highly homologous coronaviruses occur frequently under natural conditions. This implies periodic interspecific circulation of either CCoV in cats or FCoV in dogs, since mixed infections are required to give rise to recombination. It is unknown where the recombination takes place, but under experimental conditions, cats can be infected with CCoV and the possibility that FCoV might infect dogs, cannot be ruled out. Finally, recombination with an ancestral coronavirus from which FCoVs type I and Elmo/02-like CCoVs directly evolved, may not be excluded.

Replication and Cultivation

Coronaviruses enter cells via endocytosis and replicate entirely in the cytoplasm. Virus multiplication is relatively slow, as compared with a several other enveloped viruses such as orthomyxoviruses or togaviruses. The eclipse period lasts at least 6 hours, and maximum viral yield is not attained until about 24 hours after infection. Viral multiplication is optimal at 32 - 33°C and infectivity is rapidly lost at higher temperatures, or with prolonged incubation.

CCoV replicates in several types of canine and feline cell cultures; e.g., primary dog kidney, cell lines of dog thymus and skin fibroma (A72) origin, feline cell lines (CrFK, feline lung cells) and whole feline embryo cells. Cell or animal species susceptibility has recently been shown to be associated with the presence or absence of aminopeptidase-N in a species-specific manner. Cytopathic effects (cpe) are not always apparent and are poorly defined. Cpe typically appear as enlarged, bizarre shaped cells followed by focal cell detachment. Some CCoV isolates cause large syncytia in tissue culture. CCoV type I has not yet been cultivated in cell cultures.

Pathogenesis

Dogs of all ages appear to be susceptible to CCoV, however, young pups are more highly susceptible to the development of clinical infections. It is possible that all wild Canidae might be infected with the virus, but CCoV has been isolated only from coyotes with diarrhea. The scope of natural disease induced by canine coronavirus is not well understood since the pathogenesis of CCoV has not been fully defined. Early reports suggested that the infection was associated with moderate to severe gastroenteritis, but it now appears that CCoV is associated mainly with mild or subclinical disease, with low mortality rates. By extensive sequence analysis of the M and S genes of CCoV, the simultaneous presence of the two genotypes of CCoV in the feces of pups has been described. The significance of this finding is unclear since it is difficult to understand why the two CCoV genotypes occurred simultaneously in dogs. Virus in feces is the major source of infection.

The natural route of transmission is fecal-oral. Infected dogs generally shed CCoV in the feces for 6 - 9 days, but some pups have been shown to shed virus for as long as 6 months after clinical signs have ceased. CCoV is acid resistant and passes unaltered through the stomach. Viremia and generalized infection have not been observed. The surface epithelium of the small intestine is the main target of CCoV, while the colon is resistant to infection. Within 2 days after oral exposure, CCoV can be found in the upper two-thirds of the duodenal villi, which is in contrast to canine parvovirus type-2 (CPV-2) infection that destroys small intestinal crypt cells. After the fourth post-infection day there is commonly a patchy distribution of infected villi throughout the small intestine. At that time, small amounts of virus may be isolated from mesenteric lymph nodes and, occasionally, from the liver and spleen. CCoV causes a lytic infection resulting in desquamation and shortening of the villi. During this phase the virus content of feces is very high. Malabsorption and a deficiency of digestive enzymes follows, resulting in diarrhea, which can be seen 18 - 72 hours post-infection in some pups, and usually lasts for several days. Production of local antibodies (IgA) restricts spread of virus within the intestine and arrests the progress of the infection. Recovery of intestinal villi occurs in the duodenum by approximately the seventh day post-infection, and soon afterwards in the entire small intestine. Progressive healing reduces the virus shedding in the feces.

Laboratory experiments, as well as field evidence, have shown that dogs infected concurrently with both CCoV and CPV-2 leads to more severe illness than caused by either virus alone. Deaths are common during the dual infections. In a recent report, CCoV was isolated from dogs at necropsy, where this virus appeared to enhance the severity of a more enduring CPV-2 infection. In that case, CCoV was demonstrated in littermate pups that died with signs of severe hemorrhagic enteritis 15 days following recovery from CPV-2b infection.

Disease Signs

The incubation period is short. Vomiting and diarrhea may be seen by 1 - 3 days post-infection and, when clinical illness occurs, virus spreads rapidly. The virus is highly contagious and often may cause clinical signs in some dogs, with no illness occurring in other contact animals. Feces may be mucoid or watery, sometimes streaked with blood, and it is exceptionally malodorous (Fig. 2). Pups become dehydrated, depressed and anorexic, even if fluid therapy is started early. There is generally no fever, although elevated body temperatures have been observed in some cases. In contrast to CPV-2 infection, leukopenia has not been observed. Vomiting, which is much less severe than with CPV-2 infection, usually subsides after the first day of illness, but diarrhea persists several days, even for 3 - 4 weeks. Secondary infections by bacteria, parasites or other viruses, such as CPV2 or rotaviruses, may enhance and prolong the illness. However, dogs usually recover spontaneously within a week, but illness may last 2 weeks or longer. The mortality rate of CCoV infection alone is usually very low, but deaths have been reported in some kennels, especially in pups.

Coronavirus02

Figure 2. Typical stool from CCoV dog experimentally infected with CCoV (5 days post-infection). Such stools are specially malodorous.

Pathology

Morphologic lesions in CCoV infected dogs are restricted to the intestine and mesenteric lymph nodes, with dilated and enlarged intestinal loops filled with fecal material and edematous mesenteric nodes. Microscopic lesions are characterized by atrophy and fusion of intestinal villi, deepening of the crypts, and an increase in cellularity of the lamina propria with flattening of epithelial cells and discharge of goblet cells.

Immunity

Pups become immune to reinfection after recovery; however, serum antibody titers may be low. The duration of immunity is unknown. Local immunity, mediated by intestinal IgA, is considered essential for protection against CCoV infection.

Maternally transferred antibodies in dogs from affected kennels are usually low and titers are present in nursing pups for only 4 - 5 weeks.

Diagnosis

CCoV infection is difficult to distinguish clinically from enteritis caused by other agents. It is important to rule out other causes of vomiting and diarrhea such as CPV-2, enteric bacteria, parasites, poisonings and non-infectious causes of diarrhea. As with other enteric infections, diagnosis requires laboratory confirmation.

Assays that have been used for the detection of CCoV in fecal samples include electron microscopy (EM), isolation on appropriate cell cultures and polymerase chain reaction (PCR). Of the several methods used for the detection of CCoV, EM appears to be a valuable diagnostic tool. EM has been reported to be more sensitive and useful than virus isolation for detecting both coronaviruses and rotaviruses. However, the frequency of CCoV disease has probably been overestimated by diagnostic laboratories that applied EM as the principal diagnostic method. The common presence of coronavirus-like particles in feces presents difficulties in the diagnosis of CCoV by EM and requires confirmation by other tests. Immuno-electron microscopy with a specific immune serum permits confirmation of the EM diagnosis, but it requires specialized laboratories and qualified experts. Isolation in cell cultures is often used, but it is difficult and time-consuming. Commercial diagnostic laboratories have identified CCoV with certainly in only a small proportion of enteritis cases where viral isolation was the only diagnostic criterion. The virus grows on several cell lines, noted above, and it is possible to observe cpe after about 2 days of incubation. The identification of an isolate requires the neutralization of cpe with a reference sera or monoclonal antibodies.

A nested PCR (n-PCR) assay for the diagnosis of CCoV infection was recently developed. The target sequence for the amplification is a segment of the gene encoding for M protein of CCoV. The n-PCR allows the diagnosis of CCoV more rapidly than the traditional isolation on tissue cultures, and would be suitable for the diagnosis of CCoV in fecal samples when the virus is inactivated, or when the number of virions is low and cannot be detected by other tests. Recently, a taqMan® fluorogenic PCR assay was developed for the detection and quantitation of CCoV RNA in the feces of dogs. The test, which targeted the M gene is also more sensitive than conventional PCR, showing a detection limit of 10 copies of CCoV standard RNA and allowing quantitation of samples with a wide range of CCoV RNA loads. The high sensitivity, simplicity and reproducibility of the fluorogenic PCR assay makes this method especially suitable in the efficacy trials on CCoV vaccines.

The detection of CCoV antibody can be performed by virus neutralization (VN) test and ELISA. It has been demonstrated that the VN test may fail in detecting antibodies in some positive sera, providing misleading information on the epidemiological features of the infection. A recently developed ELISA was found to be more sensitive than the VN test, however, antigen prepared from CCoV infected cells can yield variable results because of differences in the care taken in preparing antigen. Recently, investigators have demonstrated that antibodies against the M protein are detected constantly in dog sera after CCoV infection. Based on those findings, an ELISA test has been developed with recombinant M protein (rMP) of CCoV type II. The test has been evaluated and may serve as an alternative diagnostic method for the detection of antibody.

The demonstrated wide diffusion of CCoV type I among dog population has underlined the need for more in-depth investigations on the serological correlations between the two genotypes. To this purpose, two recombinant polypeptides of the S glycoprotein of CCoV type I were expressed in a procariotic system and were employed to develop an ELISA for the detection of CCoV type I antibodies in dog sera. Canine sera collected from dogs vaccinated with an inactivated commercial CCoV type II vaccine and subsequently challenged with a field CCoV type II together with sera collected from dogs naturally infected with CCoV type I and negative control sera, were examined. Only the sera positive to CCoV type I reacted strongly with both the polypeptides, whereas the sera of the vaccinated dogs showed low reactivity. As CCoV type I, to date, has not been cultivated in cell cultures, the recombinant polypeptides represent a novel method to study the immunological and the pathogenetic characteristics of this new virus.

Prophylaxis and Control

Avoiding contact with infected dogs and their excretions is the only way to ensure disease prevention. Crowding, unsanitary conditions, stress during training and other environmental conditions appear to favor the development of the clinical disease. Enteric coronaviruses are moderately stable to acid and heat, but to a far less extent than are parvoviruses. CCoV is inactivated by most germicidal agents effective against other enveloped viruses. Disinfection of kennels and equipment with 3% hypochlorite solution is very effective in killing CCoV, but it does not prevent dog to dog transmission. CCoV is highly contagious; once the virus has become established in a kennel, spread of infection is difficult to control. Dogs who recover are immune to reinfection, but the duration of immunity is not known.

Vaccines

Although the value of CCoV vaccines is controversial, inactivated CCoV vaccines have been licensed in the USA. Their efficacy, however, has yet to be determined. Experiments have demonstrated limited protection against clinical disease, but not infection, at 3 weeks post-vaccination. In a recent study, the efficacy of an inactivated CCoV vaccine was evaluated in pups. In general, the results of the study have confirmed the low efficacy of the inactivated vaccine in reducing the viral shedding in feces after the CCoV challenge. Whether inactivated vaccines provide adequate immunity under field conditions is still controversial. In order to prevent infection, intraluminal antibodies (mucosal immunity) are essential since serum antibodies do not protect against infection. A modified-live (ML) CCoV vaccine licensed in the USA in 1983 was rapidly withdrawn due to a high rate (approximately 5%) of serious reactions, especially when the CCoV vaccine was given in combination with CDV and CPV-2 vaccine. Adverse reactions consisted primarily of neurologic signs and subsequent death, but some dogs had a generalized illness ("pancreatitis-meningitis-syndrome") growth retardation of pups, or sudden death. On the other hand, an ML vaccine was developed by a company in California (USA) which was claimed to be safe and effective under field conditions. However, that strain was licensed for use in dogs by another company in 1994 and post-vaccinal reactions occurred within a few weeks of introduction of that combined CCoV-CPV-CDV vaccine. A feline coronavirus vaccine was licensed in the USA, but data on that vaccine are limited.

Recently, the safety and the efficacy of a ML CCoV vaccine was evaluated in three groups of dogs: two groups were vaccinated, respectively, by the intramuscular and the oronasal routes and their responses were compared with the third group of non-vaccinated control dogs. After challenge, none of the vaccinated dogs had clinical signs. However, the dogs inoculated by the intramuscular route, as well as the control dogs, shed the challenge virus for 10 and 23 median days, respectively. In contrast, virus shedding was not observed in the dogs vaccinated by oronasal route. Even though the immune mechanisms of protection from CCoV infection are still unclear, a relationship between the levels of fecal IgAs to CCoV and the degree of protection against the challenge virus, was observed using an ELISA test. The results demonstrate a correlation between intestinal IgA antibodies and protection of dogs against infection. Assuming that IgAs against CCoV play a fundamental protective role, it seems important to evaluate vaccines in regard to both the vaccine type (ML versus inactivated) and the inoculation route.

At the present time, considering the very long viral shedding of CCoV in the feces of naturally infected pups, an effective immunization of pups against CCoV is fundamental to control virus spreading among dog populations, especially in the animal shelters. However, the 2003 AAHA Vaccine Guidelines Task Force does not recommend the use of currently available CCoV vaccines.

References
  • Decaro N, Pratelli A, Campolo M, et al. Quantitation of canine coronavirus RNA in the faeces of dogs by taqman RT-PCR. J Virol Methods 2004; 119:145-150. - Pubmed - 
  • Decaro N, Pratelli A, Tinelli A, et al. 2004. Fecal immunoglobulin A antibodies in dogs infected or vaccinated with canine coronavirus. Clin Diagn Lab Immunol 2004 Jan; 11:102-5. - Pubmed - 
  • Elia G, Decaro N, Tinelli A, et al. Evaluation of antibody response to canine coronavirus infection in dogs by western blotting analysis. New Microbiol 2002; 25:275-280. - Pubmed - 
  • Elia G, Fiermonte G, Pratelli A, et al. Recombinant M protein-based ELISA test for detection of antibodies to canine coronavirus. J Virol Methods 2003; 109:139-142. - Pubmed - 
  • Lai MMC., Holmes KV. Coronaviridae: the viruses and their replication. In: DM Knipe, P M Howley, DE Griffin, et al., eds. Fields Virology, 4th ed. Philadelphia: Lippincott Williams & Wilkins, 2001; pp.1163-1185. - Available from amazon.com - 
  • Marra MA, Jones SJM, Astell CR, et al. The genome sequenze of the SARS-asoociated coronavirus. Sciencexpress 2003 May; 300(5624):1399-1404. - Pubmed - 
  • Paul MA, Appel M, Barrett R, et al. Report of the American Animal Hospital Association (AAHA) Canine Vaccine Task Force: executive summary and 2003 canine vaccine guidelines and recommendations. J Am Anim Hosp Assoc 2003; 39(2):119-131. - Pubmed - 
  • Pratelli A, Martella V, Elia G, et al. Variation of the sequence in the gene encoding for transmembrane protein M of canine coronavirus (CCV). Mol Cell Probes 2001; 15:229-233. - Pubmed - 
  • Pratelli A, Martella V, Elia G, et al. Severe enteric disease in an animal shelter associated with dual infections by canine adenovirus type 1 and canine coronavirus. J Vet Med B Infect Dis Vet Public Health 2001; 48:385-392. - Pubmed - 
  • Pratelli A, Elia G, Martella V, et al. Prevalence of canine coronavirus antibodies in dogs in the south of Italy by an enzyme-linked immunosorbent assay. J Virol Methods 2002; 102:67-71. - Pubmed - 
  • Pratelli A, Elia G, Martella V, et al. M gene evolution of canine coronavirus in naturally infected dogs. Vet Rec 2002; 151:758-761. - Pubmed - 
  • Pratelli A, Tinelli A, Decaro N, et al. PCR assay for the detection and the identification of atypical canine coronavirus in dogs. J Virol Methods 2002; 106:209-213. - Pubmed - 
  • Pratelli A, Martella V, Pistello M, et al. Identification of coronaviruses in dogs that segregate separately from the canine coronavirus genotype. J Virol Methods 2003; 107:213-222. - Pubmed - 
  • Pratelli A, Martella V, Decaro N, et al. Genetic diversity of a canine coronavirus detected in pups with diarrhoea in Italy. J Virol Methods 2003; 110:9-17. - Pubmed - 
  • Pratelli A, Tinelli A, Decaro N, et al. Efficacy of an inactivated canine coronavirus vaccine in pups. New Microbiol 2003; 26:151-155. - Pubmed - 
  • Pratelli A, Decaro N, Tinelli A, et al. Two genotypes of canine coronavirus simultaneously detected in the fecal samples of dogs with diarrhoea. J Clin Microbiol 2004; 42:1797-1799. - Pubmed - 
  • Pratelli A, Elia G, Decaro N, et al. Cloning and expression of two fragments of the S gene of canine coronavirus type I. J Virol Methods 2004; 117:61-65. - Pubmed - 
  • Pratelli A, Tinelli A, Decaro N, et al. Safety and efficacy of a modified-live canine coronavirus vaccine in dogs. Vet Microbiol 2004; 99:43-49. - Pubmed - 
  • Risco C, Antón IM, Sune C, et al. Membrane protein molecules of transmissible gastroenteritis coronavirus also espose the carboxy-terminal region on the external surface of the virion. J Virol 1995; 69:5269-5277. - Pubmed - 
  • Wesseling JG, Vennema H, Godeke G, et al. Nucleotide sequence and expression of the spike (S) gene of canine coronavirus and comparison with the S proteins of feline and porcine coronaviruses. J Gen Virol 1994; 75:1789-1794. - Pubmed - 

 

Bill & Wendy Johnson
Mount Airy, NC 27030
(336) 786-8391

Email Us!

 

Home | The Boys | The Girls | The Nursery | Fun Pictures | Guarantee | Health