The present study evidenced that gut microbiota varies according to birth weight of puppies. These results suggest that, not only metabolic disturbances, but also intestinal dysbiosis may contribute to the higher neonatal morbidity and mortality of LBW puppies.
Firstly, significant changes in the gut microbial communities with age were observed in our study. Bacterial richness increased from D2 to D56, with the most significant increase occuring between D21 and D28. Moreover, the important shifts in the microbial composition were noticed mostly during the first days after birth. As one of the main factor shaping the gut microbiota composition and diversity, puppies’ age has already been reported in the literature, but samples were either not collected at birth or during the first week of life (D7) 16–18. Thus, our study added insights on the microbiota development, being the first to give a clear picture of the microbiota development on the first days after birth (D0, D2 and D7).
When looking at beta diversity, two main microbial profiles were observed. The first one corresponded to the microbiota profile at a very young age, around the first week of life, when it is highly dynamic and unstable, with the largest inter-individual composition observed. At that time, environmental conditions are favourable for aerobes and facultative anaerobes (Moraxellaceae, Clostridiaceae, Aerococcaceae, Enterobacteriaceae). The second profile corresponded to a more mature and anaerobic GIT with samples from D21 to D56, mainly populated by bacteria such as Bacteroidaceae, Selenomonadaceae and Succinivibrionaceae. The dissimilarity between those two clusters could be drafted by nutritional shifts taking place in puppies during the first weeks of life. In our study, puppies were fed milk exclusively until 14 days of life, whereas older puppies had access to solid food (kibbles). Previous studies demonstrated that the transition from milk to a solid diet is usually followed by an increase in bacterial diversity 16,19,20. Moreover, dogs fed a high carbohydrate diet (kibbles) present a high Bacteroidetes/Firmicutes ratio 21,22, also observed in our study after D21. An increase of the Bacteroidetes phylum, and particulary Prevotella and Bacteroides genera, as noted in our population on day 21, is known to produce important short-chain fatty acids from carbohydrates and glycans 23,24. Short-chain fatty acids play a major role in the maintenance of gut and immune homeostasis, which could also explain the second cluster at D21.
In addition to nutritional transition, the two main bacterial profiles could also be explained by physiological shifts occuring in the GIT. Indeed, it has been shown in most mammals that at birth the GIT is filled with oxygen. The maturation of the GIT results in the establishment of an anaerobe environment, thanks to the concomitant action of oxygen-consuming bacteria (i.e., facultative anaerobes) and of colonocyte metabolism, and in particular the beta-oxydation of fatty acids 25. At the beginning, the positive redox potential in the gut makes the perfect environment for strict and facultative aerobic bacteria (mainly belonging to Proteobacteria and Firmicutes phyla) to settle 26–28. As those bacteria and the colonocytes consume oxygen, strict aerobic bacteria are quickly replaced with opportunistic aerobic facultative bacteria which become the dominant taxa on just a few days of age 27, as observed in puppies since D2 and until D21 in this study. Those phenomena result in the reduction of the redox potential of the gut, and of better physicochemical conditions for the establishment and growth of obligate anaerobic bacteria. This neocolonization of the gut explains the increase of bacterial richness over time, together with the enrichment of food sources. In our results, Moraxellaceae and Aerococcaceae, two aerobic strict families, represented around 25% of the relative abundance of puppies’ microbiota at birth. At D2, they already almost disappeared, while Enterobacteriaceae alone, mostly comprised of facultative anaerobic bacteria, such as E. coli, represented almost 30% of the relative abundance, inducing the observed decrease in bacterial richness. On later days, the abundance of obligate anaerobes increased, such as Succinivibrionaceae, Bacteroidaceae and Bifidobacteriaceae, while the abundance of facultative aerobes decreased, making our results consistent with the literature in other mammalian species (Fig. 11) 28–30.
For the first time, some important differences in the gut microbiota development in LBW puppies versus NBW were demonstrated in our study, most probably related to the level of oxygen in the GIT and the immaturity of colonocyte maybe more pronounced in those particular puppies. Indeed, despite no significant differences, we observed that LBW puppies had lower abundances of Aerococcaceae and Moraxellaceae at birth. The lack of those strict aerobic bacteria might have prevented an adequate consumption of oxygen in the GIT during the first days of life. The higher level of oxygen might have induced an earlier and increased colonization of facultative anaerobes, which strived in such an environment compared to the GIT of NBW puppies. The reduced consumption of oxygen also delayed the colonization by strict anaerobes bacteria and subsequently the proper setting of the immune system of newborns 31,32. Indeed, this abnormal dominance of facultative anaerobic bacteria have been widely reported in preterm infants and is known to induce alteration of the intestinal barrier and immunological functions of the host 31,33,34.
Enterobacteriaceae, Clostridiaceae and Lachnospiraceae observed in our study at higher abundance at day 2 and 21 in LBW compared with other groups, are known to be opportunistic bacteria leading to favourable conditions for diseases and specifically inflammatory bowel diseases in the canine species 35–37. Among those families, we found higher abundances of E. coli, Clostridium perfringens and Tyzzerela in LBW puppies. E. coli is demonstrated to be involved in intestinal diseases and systemic infections in newborn puppies 38, even though this commensal species remains present in high quantities in healthy puppies compared to adult dogs 39. In many animal species, Tyzzerela bacteria are causative agents of Tyzzer’s disease, a usually fatal infectious disease characterized by diarrhea, abdominal distention and hepatic necrotic lesions 40,41, already described in puppies 42,43. As C. perfringens, this species has also been observed in higher abundances in preterm infants and piglets as well, linked to higher risk of necrotizing enterocolitis 32,44.
These findings suggest that LBW puppies develop altered gut microbiota during the first days of life, being most probably associated with their higher risk of death. As observed through their significantly lower APGAR score, LBW puppies are weaker than NBW at birth, making them struggle to reach the mammary glands of their mother and consume colostrum 45. This could explain the initial differences of bacterial composition during the first days of life between LBW and NBW. It remains delicate to determine if the composition of the microbiota induces the weakness of the LBW puppies or the opposite, but those results bring new crucial information on the understanding of the higher risk of mortality of LBW puppies 46.
Another relevant genus of interest observed in lower abundances in LBW puppies in the present study was Phascolarctobacterium, a strict anaerobe which uses succinate to produce acetate and propionate. This genus has been associated with energy metabolism regulation in dogs and humans 47–49. A recent study highlighted that LBW puppies were at higher risk of getting overweight once adult 50, thus, the lower abundance of Phascolarctobacterium in LBW puppies could be an interesting biomarker to follow.
Finally, an important result from this study is the absence of observed differences in the microbiota composition between LBW and NBW puppies after the third week of life. All 57 puppies studied were healthy and did not present any diarrheic symptoms during the first two months of life. This suggests that LBW puppies strong enough to survive the delayed microbiota development during the neonatal period were able to catch up with other puppies and harbour a “normal” microbiota after a month. This is coherent with previous results highlighting that most neonatal deaths occurred during the first weeks of life, with very-low-birth weight ones dying earlier than NBW ones 1,46.
The results of the present study suggest potential preventive strategies to reduce the mortality of LBW puppies. For example, specific pro and prebiotic strains could be given to LBW puppies as soon as possible to stabilize their microbiota and limit potential dysbiosis. Also, avoiding artificial milk and antibiotic as much as possible is highly recommended, as colostrum is a good source of bacteria for the establishment of the microbiota and antibiotic are known to induce dysbiosis, with in particular the proliferation of facultative anaerobic bacteria 51,52.
Differences between HBW and NBW were more limited than LBW, with very few differences observed and not recurring from one day to another. The most notable differences were observed at D2 and D21 with higher relative abundances of Faecalibacterium and Bacteroides respectively. The role of the butyrate-producing Faecalibacterium in gut health has been widely studied recently, to the point it became a bioindicator of inflammatory bowel diseases when its abundance decreases, both in humans and dogs 53,54. Interestingly, this genus seems to be in lower abundance in young puppies compared to adult dogs, so an increase in the abundance of this genus can be linked to the maturity of the puppy 39. Another genus observed with higher abundance in HBW puppies in this study, Bacteroides, is shown to be present in lower abundance in preterm infants 32. This genus also uses glycans to produce butyrate, reinforces the protection from pathogens in the gut, and as obligate anaerobe, is synonym of the proper maturation of the GIT of the host 55,56. Thus, the higher abundances of Faecalibacterium and Bacteroides in HBW puppies during the neonatal period could suggest the maturation of their microbiota happened earlier than in the other puppies. At birth, the HBW puppies might be able to reach the mammary gland with more ease compared to their brethren, allowing them to get more colostrum and potentially, acquiring a more complete microbiota from it 5,57. At three weeks old, when the first dry diets were presented to puppies, it might be suggested that HBW ingested more food to support the higher needs of their body weight, inducing the increased abundance of Bacteroides linked to dry food consumption 21. Those results would suggest that the early maturation of the GIT microbiota of HBW puppies would give them a better protection against sepsis and other gastrointestinal diseases during the first weeks of life 58. Since HBW puppies are not specifically studied but usually considered among NBW, it remains difficult to compare those results with literature 59. If this hypothesis of an early maturation proved to be true, then fecal microbiota transplantation from HBW puppies to LBW puppies could prove to be another effective strategy to stabilize and improve the microbiota stability of LBW puppies using the more mature one of the HBW.
Some limits have to be considered when looking at the results of the present study. First, only 57 puppies were involved, of which 14 were considered LBW and 14 were HBW. While this remains higher than similar studies conducted on other species 32,60, individual variability of the microbiota composition is a major challenge. On top of it, the LBW studied were in good health and survived the first two months of life, meaning their microbiota might not be representative of LBW puppies dying during the neonatal period. All 57 puppies were from the same kennel, which allowed to reduce variability between individuals (same food, same environment), but limit the extrapolation to the whole canine species. Also, previous results highlighted that brethren pups have a closer microbiota composition compared to unrelated ones, meaning the litter effect might have an impact on the results 61. However, since puppies from a same litter ended up in different quartiles, it allowed to reduce the impact of individual variability of the microbiota composition and confirmed the differences highlighted were most likely a consequence of the birth weight.
Nevertheless, while the impact of LBW on puppies’ health had already been studied 3,45,50, this study is the first one to date to describe differences in the fecal microbial populations based on the birth weight of puppies. Further studies, comparing the microbial profile in LWB puppies dying during the first weeks of life with the surviving ones would be desired to identify bacteria responsible for neonatal mortality in the canine species. Another interesting use of these data would be to develop early life microbial biomarkers per birth weight category to predict the risk of diseases later in life, and to propose therapeutical or nutraceutical strategies to orientate the microbial trajectory.