Anne McCartney, Department of Food and Nutrition Sciences, University of Reading, Reading RG6 6AP, UK. Tel.: +44 (0)118 3788593; fax: +44 (0)118 3787708; e-mail: email@example.com
Weaning is a stressful process for kittens and is often associated with diarrhoea and the onset of infectious diseases. The gastrointestinal (GI) microbiota plays an essential role in host well-being, including improving homoeostasis. Composition of the GI microbiota of young cats is poorly understood and the impact of diet on the kitten microbiota unknown. The aims of this study were to monitor the faecal microbiota of kittens and determine the effect(s) of diet on its composition. Bacterial succession was monitored in two groups of kittens (at 4 and 6 weeks, and 4 and 9 months of age) fed different foods. Age-related microbial changes revealed significantly different counts of total bacteria, lactic acid bacteria, Desulfovibrionales, Clostridium cluster IX and Bacteroidetes between 4-week- and 9-month-old kittens. Diet-associated differences in the faecal microbiota of the two feeding groups were evident. In general, fluorescence in situ hybridization analysis demonstrated bifidobacteria, Atopobium group, Clostridium cluster XIV and lactic acid bacteria were dominant in kittens. Denaturing gradient gel electrophoresis profiling showed highly complex and diverse faecal microbiotas for kittens, with age- and/or food-related changes seen in relation to species richness and similarity indices. Four-week-old kittens harboured more diverse and variable profiles than those of weaned kittens.
Over the past 10 years, an increasing number of studies have investigated the predominant gastrointestinal (GI) and faecal microbiotas of healthy adult cats (Papasouliotis et al., 1998; Sparkes et al., 1998; Marshall-Jones et al., 2006; Ritchie et al., 2008; Lubbs et al., 2009; Handl et al., 2011). However, very little information is available on the microbiota of kittens. Establishment of the gut microbiota in neonates is considered to affect postnatal development of the immune system and long-term health (Bailey et al., 2005). One early study examined the GI microbiota of suckling kittens (1–3 weeks’ old) and adult cats fed either a standard laboratory feline ration or a chemically defined liquid ration (Osbaldiston & Stowe, 1971). Enterococcus, Escherichia coli and Lactobacillus were shown to be the predominant bacteria both in kittens and adult cats.
Age-associated changes in the feline faecal microbiota were mentioned in a study by Patil et al. (2000) investigating the microbiota of cats from three age groups (young, adult and elderly). Bifidobacterial levels were significantly lower in the elderly group compared with the other two groups, while Clostridium perfringens counts were lower for young cats than the other two groups.
Previous studies have indicated that intestinal bacterial succession in young animals (humans, chicks, pigs and calves) follows a similar pattern during the first few weeks of life (Mackie et al., 1999). Coliforms and streptococci dominate the microbiota within a few days of birth, obligate anaerobes appear later, and clostridia and lactobacilli may also be present in most hosts within a short period of time. It is assumed that a similar trend is true in companion animals (i.e. dogs and cats). Indeed, Buddington (2003) demonstrated that the GI tract of neonatal dogs was rapidly colonized by bacteria from the birth canal and the surrounding environment, and an age-related increase in anaerobic bacterial proportion was found. Large-scale bacterial changes associated with age were shown in the canine GI tract, particularly during weaning (Buddington, 2003).
Diet is essentially important to the neonate, providing nutritional requirements and aiding development/maturation of the immunological system and other organs. Diet has also been shown to affect bacterial succession in human infants (Favier et al., 2002; Hopkins et al., 2005; Roger & McCartney, 2010). Favier et al. (2002) found significantly more bifidobacteria and less-complex microbial communities in breast-fed babies compared with formula-fed babies during the first few days of life. Furthermore, the introduction of solid food coincided with major shifts in the microbial profiles of the breast-fed infant, as did the subsequent withdrawal of breast milk from the diet. Commercially available milk replacer has been reported to support normal kitten growth rate, but results in diarrhoea and cataract formation compared with queen's milk (Remillard et al., 1993) – with cataract formation related to deficiency of certain amino acids. Weaning is generally considered to be stressful for kittens and outbreaks of diarrhoea and disease are common (Kirk et al., 2000). The objectives of this study were to investigate the predominant faecal microbiota of kittens, monitor the bacterial succession over time and examine the effects of two commercial kitten foods on the feline faecal microbiota.
Materials and methods
Unless otherwise stated, all chemicals were supplied by Sigma-Aldrich (UK).
A longitudinal study was performed on 12 kittens and their litters to examine bacterial succession in faecal microbiota, and compare Hill's Science Diet® Kitten Healthy Development Original to the best selling commercial kitten foods in USA at the time of the study (namely, Purina® Kitten Chow®). Group A received Purina® Kitten Chow® and group B received Hill's Science Diet® Kitten Healthy Development Original (Table 1). Both foods are complete balanced foods for growth and lactation. Dams were fed the same foods (i.e. Purina® Kitten Chow® or Science Diet® Kitten Healthy Development Original) as their kittens.
Table 1. Composition of the kitten foods (% dry matter unless stated otherwise)a
The study protocol was reviewed and approved by the Institutional Animal Care and Use Committee. All cats were immunized against feline panleukopenia, calici and herpes viruses and rabies, and none had chronic systemic disease on the basis of results of physical examination, complete blood count determination, serum biochemical analyses and urinalysis, and were negative for faecal examination for parasites. The kittens were housed with their mother and littermates during the suckling period (birth to 8 weeks of age). Kittens were weaned from their dams at 8 weeks of age, with transition from mother's milk to food between 7 and 8 weeks of age (i.e. kittens received dry food as a choice while they were still nursing; 7–8 weeks of age). After weaning, kittens were house together according to dietary group in separate spacious rooms with natural light [group A (n =6) in one room and group B (n =6) in another]. Food was continuously available throughout the day until their daily caloric requirements were consumed and cats had free access to water. Cats received routine grooming and experienced behavioural enrichment through interactions with each other, by daily interaction and playtime with caretakers and access to toys. All animals were cared for in accordance with the Hill's Institutional Animal Care and Use Committee protocols.
Faecal samples were collected from the kittens’ litters at 4 (n =9) and 6 weeks of age (n =14) and from each of the 12 kittens at 4 and 9 months of age. Litter boxes were provided in each room and technicians closely monitored these on sampling days. Stools were collected following defecation (within 15 min for litter samples and immediately after for weaned kittens), immediately stored in anaerobic pouches [labelled according to litter sample or kitten name (weaned samples)] and frozen at −20 °C. It was not possible to identify which kitten within the litter had produced each sample (preweaning). In total, 23 faecal samples were collected from the litters and 24 faecal samples postweaning.
Analysis of the faecal microbiota
Blind-coded faecal samples were shipped on dry ice to The University of Reading. The faecal samples were held at −20 °C before microbial analysis. Faecal samples were thawed and c. 2 g diluted [1 : 10 (w/w)] in sterile phosphate-buffered saline (PBS: 0.1 M, pH 7.4; Oxoid, UK) and homogenized by shaking with glass beads for at least 1 min. Each faecal homogenate was then processed for FISH and denaturing gradient gel electrophoresis (DGGE) analyses.
Fluorescence in situ hybridization
Samples were processed for FISH as described previously (Jia et al., 2010). Bacterial group-specific and genus-specific 16S rRNA and 23S rRNA gene-targeted oligonucleotide probes were commercially synthesized and monolabelled (at the 5′ end) with the fluorescent dye Cy3 (SigmaGenosys, UK). Details for the probes used in this study can be found in Table 2. The nucleic acid stain 4′,6-diamidino-2-phenylindole (DAPI) was used for enumerating total bacteria. FISH was performed as described by Martín-Peláez et al. (2008). Bacteria and total bacteria were enumerated as follows:
where 8 is the dilution factor for the faecal sample; Y is the average cell count per field of view; 6732.42 is the magnification constant (number of fields of view per well); 50 is the dilution factor for the FISH sample in each well; and Z is the dilution factor for the specific probe/DAPI stain being examined.
Table 2. Details for oligonucleotide probes used in this study
Samples were processed for DGGE as described previously (Jia et al., 2010). PCR-DGGE was performed using the universal primers P2 (5′-ATTACCGCGGCTGCTGG-3′) and P3 (5′-CGCCCGCCGCGCGCGGCGGGCGGGGCGGGGGCACGGGGGGCCTACGGGAGGCAGCAG-3′), which amplify a 193-bp fragment of the V3 region of the 16S rRNA gene (between E. coli positions 341 and 534) (Muyzer et al., 1993).
A general linear model univariate analysis was performed in spss 14.0 (SPSS Inc.) to analyse the effects of age, group and the interaction of age × group on different bacterial groups. When a statistical significance was detected, Tukey post hoc test for multiple comparisons was used to identify the specific differences. Statistical significance was accepted at P <0.05.
DGGE banding patterns were analysed with totallab tl120 software (version v2006f; NonLinear Dynamics Ltd., UK) and imported to totallab tl120dm (version v2006f; NonLinear Dynamics Ltd.) for comparisons across different gels. The number of bands was calculated to indicate species richness. Cluster analysis and calculation of similarity indices between DGGE profiles were performed with Pearson correlation coefficient and the unweighted pair group method with arithmetic mean (UPGMA) dendrogram type. totallab tl120dm does not afford bootstrapping of dendrograms, so validation of the observed clusters in the dendogram was performed by comparing the resemblance values of the similarity matrix with those on the dendogram (van Verseveld & Röling, 2004). Shannon Weaver diversity index was calculated according to van Verseveld & Röling (2004), to access the diversity of the faecal microbiota profiles. Statistical differences in number of bands, similarity values and diversity were analysed in the same model as for the bacterial groups (i.e. univariate analysis and Tukey post hoc test). P <0.05 was considered statistically significant.
Composition of the faecal microbiota of kittens
Bif164 levels were significantly higher in group A at 4 months of age than was seen for their litters at 4 weeks of age (Table 3). Bif164 levels in group B were significantly lower at 9 months of age than were seen for their litters at 4 weeks of age. Furthermore, group B had significantly lower Bif164 counts compared with group A at 9 months of age (Table 3). An interaction of age and diet was also shown for CFB719, with significantly higher counts seen in 9-month-old group A kittens compared with their litters at 4 weeks of age.
Table 3. Investigation of the effects of age and food type on the faecal microbiota of kittensa
Age × diet P-value
Group A (n =5)
Group B (n =4)
Group A (n =10)
Group B (n =4)
Group A (n =6)
Group B (n =6)
Group A (n =6)
Group B (n =6)
Data are expressed as mean values ± SD [log10cells per gram faeces (wet weight)].
Significantly different to Bif164 levels for same group at 4 weeks of age (P <0.05).
Significantly different to Bif164 levels of group A at 9 months of age (P <0.05).
Significantly different to CFB719 levels of group A at 4 weeks of age (P <0.05).
An age-related effect was observed for DAPI, Lab158, DSV687, Erec482 and CFB719 counts. Preweaned kittens (both 4- and 6-week-old litter samples) had significantly higher DAPI and significantly lower CFB719 counts than kitten samples collected at 9 months of age (P <0.01 and P <0.05, respectively). Furthermore, litter samples from 4-week-old kittens had significantly higher DSV687 levels than weaned kittens (both 4- and 9-month-old kittens; P <0.01). Litter samples from 4-week-old kittens had significantly lower Lab158 counts compared with litter samples from 6-week-old and weaned kittens (P <0.001 and P <0.01, respectively). In addition, litter samples from 6-week-old kittens had significantly higher Prop853 counts than 4-month-old kittens (P <0.01) and 4-month-old kittens had significantly higher Lab158 levels than 9-month-old kittens (P <0.05).
Bacterial composition was also affected by food type, with significantly lower Bif164 and Erec482 counts and significantly higher Chis150 counts seen for group B compared with group A (P <0.05).
Overall, the relative predominance (per cent of total bacteria) of each bacterial group examined was similar between the two groups of kittens, with the exception of Bif164 postweaning (Supporting Information, Table S1). Prior to weaning, Bif164 counts comprised <12% of the DAPI counts of all litter samples (range 0.01–11.12% of DAPI). However, at 4 months of age Bif164 counts made up 6.8–53.4% of the DAPI counts of group A (average, 27%) and 1.1–6.4% for group B (average, 4%). This then diminished in both groups by 9 months of age, with 0.5–33.6% and 0.3–1.1% of DAPI counts accounted for by Bif164 (averages, 18% and 0.6%) in groups A and B respectively. Although the Erec482 proportions also differed between the two kitten groups at 4 weeks of age (c. 12% for group A and c. 2% for group B), thereafter Erec482 proportions were relatively similar between the two groups (c. 9%, 9% and c. 13% for all samples from kittens aged 6 weeks, 4 months and 9 months, respectively). In general, Ato291 counts formed the predominant bacterial group throughout the study (averaging between 10% and 23%; Table S1), although Bif164 was co-dominant in group A postweaning. The remaining bacterial groups each made up <1% of the kitten microbiota, with the exception of Lab158 – which generally comprised <5% of DAPI counts. However, two 4-month-old kittens (one from each group) harboured Lab 158 populations >15%; three other 4-month-old group B kittens and one 4-month-old group A kitten displayed 5–10% Lab158.
Biodiversity of the faecal microbiota of kittens
The universal DGGE profiles obtained for the faecal samples collected from kittens, whether litter samples or samples from individual kittens, demonstrated a complex microbiota (with >20 bands per profile) (Figs 1 and 2). Some common bands were observed (e.g. band A, Fig. 1; band D, Fig. 2), as well as bands specific for certain age groups (irrespective of group). For example, one band was not seen in litter samples taken when the kittens were 4 weeks old but was evident in all other profiles (band C, Fig. 1). Another band was only found in 4-month-old kittens (5/6 group A profiles and 4/6 group B profiles; band F, Fig. 2). This demonstrates that certain bacterial genotypes are predominant in the kitten faecal microbiota while others fluctuate. Interestingly, species richness was seen to decrease after weaning, with significantly more bands per profile for 4-week-old litter samples (30 ± 5) than seen for 4- (24 ± 5) and 9-month-old kittens (25 ± 4) (P <0.05). Diversity was also seen to be higher for 4-week-old litter samples (Shannon Weaver diversity index; Table 4), with significantly higher diversity than that seen for 4-month-old kittens (P <0.05).
Table 4. Investigation of the effects of age and food type on the diversity of the DGGE profiles of kittens’ faecal microbiotab
An age-related effect was also shown in the Pearson similarity analysis, with significantly lower similarity observed between litter samples from 4-week-old kittens compared with between litter samples from 6-week-old kittens and faecal samples from 4- and 9-month-old kittens (P <0.001). Age × diet interaction indicated that DGGE profiles of 4-month-old kittens were more similar to each other for group A than group B (Table 4); however, the opposite was seen for 9-month-old kittens, with significantly higher similarity between group B profiles than for group A profiles. A diet-related effect was also shown, with significantly higher similarities between the profiles of group A samples across all ages than seen for group B samples (P <0.01).
Comparison of profiles using the UPGMA algorithm generally demonstrated clustering relative to age and diet (Fig. 3). Profiles from litter samples collected when the kittens were 4 weeks old did not cluster together as well as the profiles from weaned kittens (4 and 9 months of age), with lower similarity (%) observed (Table 4). In addition, profiles from group A generally clustered better than those from group B (especially the litter samples from 6-week-old kittens), although the profile for kitten A3 at 9 months of age clustered closer to the profiles obtained for group B kittens at the same age than to the profiles from other group A 9 month-old kittens.
Compared with numerous studies on the GI microbiota of human infants (e.g. Favier et al., 2002; Heilig et al., 2002; Hopkins et al., 2005; Palmer et al., 2007; Roger & McCartney, 2010), there are rather limited data available on the microbiota of kittens (Osbaldiston & Stowe, 1971; Vester et al., 2009). In this study, we monitored the development of the kitten faecal microbiota. Ato291 counts were highest for kittens in this study and were relatively stable throughout. Probe Ato291 detects some, but not all, members of the family Coriobacteriaceae (namely Cryptobacterium curtum and Eggerthella, Collinsella, Atopobium and Olsenella spp.). The creation of the family came with the description of the genus Atopobium, which was described in 1992 to accommodate the reclassification of Lactobacillus minutus (Atopobium minutum), Lactobacillus rimae (Atopobium rimae) and Streptococcus parvulus (Atopobium parvulum) (Collins & Wallbanks, 1992). Collinsella aerofaciens and Eggerthella lenta are readily isolated from human faeces and novel equol-degrading members of the genus Eggerthella have been described recently. However, little information is available on the role(s) and importance of these bacteria in the GI tract of humans and other animals. Previous studies investigating the faecal microbiota of cats have not examined this group of bacteria. However, studies carried out in this laboratory have consistently found this bacterial group to be predominant in the feline faecal microbiota (including healthy senior cats and cats with inflammatory bowel disease) (Abecia et al., 2010; Jia, 2009).
High Erec482 counts were also found from the faecal samples of kittens. This is in accordance with the findings of Terada et al. (1993) in healthy feline faecal samples and our group in other feline studies. Furthermore, Ritchie et al. (2008) suggested that Clostridium and Eubacterium were predominant members of the feline GI tract, having found 54% of the clones from clone libraries obtained from feline stomach, duodenum, jejunum, ileum and colon samples belonged to the order Clostridiales (with members of the Clostridium clusters XIVa and IV the most common). Recent pyrosequencing data from the same research group found 65.14 ± 23.11% of bacterial sequences from feline faeces belonged to the class Clostridia (Handl et al., 2011). Caution should be taken, however, when extrapolating predominance based on cloning studies, as highlighted by the relative lack of Bifidobacterium sequences in clone libraries constructed using universal bacterial primers despite Bifidobacterium-specific primers suggesting 100% of cats harboured bifidobacteria (Ritchie et al., 2010). Using FISH, bifidobacteria have been shown to be one of the predominant members of the feline faecal microbiota (Abecia et al., 2010), including kittens.
The quantitative data herein confirms the findings of Handl et al. (2011) that Clostridium cluster XIVa (detected by Erec482) is the most prevalent Clostridium cluster in the feline faecal microbiota, although the numerically predominant members of the kitten faecal microbiota were shown to be Coriobacteriaceae (detected by Ato291). Interestingly, Actinobacteria was the second most abundant phylum found by Handl et al. (2011), with the majority of OTUs in cats belonging to the family Coriobacteriaceae (notably Eggerthella and Olsenella, both of which are detected by the probe Ato291).
Food containing higher carbohydrate content, lower protein and lower fat induced significantly higher Bif164 and Erec482 counts and significantly lower Chis150 counts in weaned kittens, compared with the higher protein and higher fat kitten food. This is consistent with previous work by Lubbs et al. (2009), who showed that the protein content of food affects the faecal microbiota of cats. They reported significantly higher bifidobacteria and significantly lower C. perfringens levels in cats fed moderate protein (MP, 34.34% of dry matter) compared with cats on high protein (HP, 52.88% of dry matter) diets. The same diets were later tested on kittens from 4 weeks of age until 16 weeks of age, with significantly higher bifidobacteria and lactobacilli found in the kittens fed MP compared with those fed HP food (Vester et al., 2009). However, significantly higher levels of E. coli and C. perfringens were also seen in the kittens fed MP compared with those fed HP diets.
Lactic acid bacteria (as detected by probe Lab158) was another bacterial group of particular interest in our study, as members of this group have been recognized as candidate probiotics for companion animals (Benyacoub et al., 2003; Marshall-Jones et al., 2006; Biagi et al., 2007). Early data reported that Lactobacillus was predominant in the stomach and jejunal contents and commonly isolated from mid-colon contents of kittens (1–3 weeks of age) (Osbaldiston & Stowe, 1971). Age-related changes in the faecal Lab158 counts were seen in the present study, forming one of the dominant bacterial groups in some 4-month-old kittens. Few data are available on cats in early age; however, Lactobacillus was found in high numbers in the faeces of healthy adult cats (Sparkes et al., 1998; Lubbs et al., 2009) and observed in 92% of pet cats using Lactobacillus-specific primers (Ritchie et al., 2010).
DGGE has been effectively applied to monitor the succession of the intestinal microbiota and possible shifts caused by diet in humans and other animals (Simpson et al., 1999, 2000; Favier et al., 2002; Roger & McCartney, 2010). In this study, DGGE profiles demonstrated that each kitten harboured a complex microbiota. Notably, 4-week-old kittens harboured a more diverse microbiota than weaned kittens (4 and 9 months of age). The introduction of food to the diet, and its composition, had an important effect on the development of the kitten microbiota. DGGE profiles of weaned kittens clustered according to food, with greater similarity seen between profiles for kittens fed the higher carbohydrate, lower protein and fat food than for those of the kittens on the higher protein and fat food. This follows the same trend as seen for piglets (Konstantinov et al., 2003), with dramatic changes in the microbiota shown after weaning and higher bacterial diversity and quicker stabilization of the microbiota postweaning associated with nondigestible fermentable carbohydrates (beet pulp and/or fructooligosaccharides).
In conclusion, this study has described the quantitative and qualitative development of the faecal microbiota of kittens from 4 weeks to 9 months of age and the impact of two kitten foods on the faecal microbiota. Notable changes in the bacterial populations (e.g. DAPI, Lab158, DSV687 and CFB719 counts), together with alternations in the DGGE profiles, were most significant between 4-week-old and 9-month-old kittens. Food was shown to affect the development of the kitten microbiota. More work is required to elucidate the role(s) of different bacterial groups (and the microbiota as a whole) in the feline GI tract. For example, what is the effect of the noted differences in the feline microbiota of kittens fed the higher carbohydrate, lower protein and fat diet from those fed the higher protein and fat diet on GI function (particularly in relation to gut and animal health)?
This work was supported by grants from Hill's Pet Nutrition, Topeka, KS, which requested publication of this work. Paul Chatfield (University of Reading) is thanked for assistance with statistical analyses.