Selective inbreeding does not increase gut microbiota similarity in BALB/c mice

Abstract

Inflammatory diseases in mouse models are under strong impact from the gut microbiota. Therefore increased interindividual gut microbiota similarity may be seen as a way to reduce group sizes in mouse experiments. The composition of the gut microbiota is to a high extent defined by genetics, and it is known that selecting siblings as mothers even in inbred colonies may increase the gut microbiota similarity among the mice with 3–4%. We therefore hypothesized that selective breeding of mice aiming at a high similarity in the gut microbiota would increase the interindividual similarity of the gut microbiota. BALB/cCrl mice were, however, found to have a mean heterozygosity of only 0.8% in their genome, and selection of breeders with a high similarity in the gut microbiota for three generations did not change the overall gut microbiota similarity, which was 66% in the P generation and 66%, 64% and 63% in the F1, F2 and F3 generations, respectively. Increased gut microbiota similarity in closely related mice in inbred mouse colonies is, therefore, more likely to be caused by other factors, such as imprinting or different intrauterine conditions, rather than by residual heterozygosity.

Inflammatory diseases in mouse models are under strong impact from the gut microbiota. ¹ Therefore, an increased interindividual similarity may lead to reduced variation in parameters monitored in studies of inflammatory diseases, such as cytokines, ² glucose tolerance ³ and diabetes incidence, ⁴ and therefore, it may be seen as a tool to reduce group sizes in such studies. ⁵ The composition of the gut microbiota is to a high extent defined by genetics. ⁶ Outbred colonies of mice have an interindividual gut microbiota similarity of approximately 67%, while it is approximately 77% in inbred colonies, ⁷ in which it also differs strictly between strains. ⁸ Selecting siblings as mothers may increase the similarity with 3–4%. ⁹ Inbred mice are claimed to have an inbreeding coefficient >98%, ¹⁰ although concrete studies on this matter are few. ¹¹ On the other hand, separation of rodent strains into separated subcolonies gives rise to phenotypic differences. Selective inbreeding may therefore be hypothesized to be a way to increase the interindividual gut microbiota similarity in a colony of mice. Alternatively, factors such as imprinting or intermaternal differences in uterine conditions may explain some of the observations described above. ^12–14

Six female and seven male BALB/cCrl mice (Charles River, Sulzfeld, Germany) (P generation) had at the age of five weeks, at which age the gut microbiota seems to have established itself in a more permanent form, ¹⁵ under anaesthesia with Hypnorm (VetaPharma, Leeds, UK) and Dormicum (Roche A/S, Hvidovre, Denmark) approximately 5 mm of the tail tip as well as caecum contents sampled as previously described ¹⁶ according to approval from the Danish Animal Experimentation Board. After surgery analgesic (Rimadyl^®; Orion Pharma, Nivå, Denmark, 5 mg/kg subcutaneously) was given. Water with 1.2 mL of Ampivet^® (Ampicillin; Boehringer Ingelheim, Copenhagen, Denmark) in 300 mL was provided for seven days starting from the day of surgery. Mice were housed in standard cages (type 1290; Tecniplast, Buggugiate, Italy) on aspen bedding (Tapvei, Tallinn, Estonia) with supplement of Enviro-dri and Alpha-Nest nesting material (SSP, Milford, NJ, USA), Shepherd's Shacks (regular, SSP) and an aspen chew block (Tapvei) under a 07:00–19:00 h light/dark cycle and 440 lx light intensity in front of the cage rack. Altromin 1314 fortified diet (Altromin GmbH, Lage, Germany) was fed ad libitum. Relative humidity was 55 ± 10%, and the temperature was 20–24°C. Breeding pairs and litters were separated at four weeks of age. The 13 tail tips of the P generation were genotyped using the single nucleotide polymorphism (SNP) chip ‘Affymetrix^® Mouse Diversity Genotyping Array’ (MouseDIVm520650) (Affymetryx, High Wycombe, UK). The gut microbiota of the 13 P generation caecum contents were characterized by the polymerase chain reaction and denaturing gradient gel electrophoresis (DGGE) and subjected to cluster analysis using the Dice similarity coefficient with a position tolerance setting of 1% and 1% position tolerance for band composition as previously described. ⁹ The two P generation pairs with the highest similarity in gut microbiota as based on cluster analysis were selected for breeding, and the offspring (F1) caecal microbiota was characterized by DGGE at the age of five weeks. The breeding pair with the highest gut microbiota similarity was then selected separately within each line and this was repeated for F2, while for F3 the caecal microbiota was characterized without further selection for inbreeding. One litter was born within each line of F1 and F2, while two litters per line were born within F3.

There was a mean heterozygosity of 0.8% in the genome of the P generation, which had an overall similarity in their gut microbiota of 66%. There was no correlation between SNP clustering and DGGE clustering when tested by chi²-test on clusters. Two breeding pairs were found with a gut microbiota similarity of 78%. In F1, F2 and F3 the overall similarity of the gut microbiota was 66%, 64% and 63%, respectively, while the highest gut microbiota similarity of the breeding pairs was 79% in both the F1 and F2 generations, respectively (Table 1).

Table 1

Gut microbiota similarities of BALB/cCrl mice subjected to selective breeding aimed at increasing these similarities

	Line 1			Line 2			Overall
Generation	Male/female GM similarity in selected breeding pair %	Litter size	GM similarity in litter %	Male/female GM similarity in selected breeding pair %	Litter size	GM similarity in litter %	Male/female GM similarity in selected breeding pair %	Total number	GM similarity %
P							78	13	66
F1	78	4	73	79	7	67		11	66
F2	79	5	66	66	6	62		11	64
F3		4	67		6	67		18	63
		4	67		4	55

Lines 1 and 2 (F1, F2 and F3) are based upon two breeding pairs selected from one parent breeding pair (P)

GM: gut microbiota; P: parent generation; F1, F2 and F3: generations 1, 2 and 3

In conclusion, the overall similarity of the gut microbiota was quite stable over generations. On the positive side is that these commercially available inbred mice had a low mean heterozygosity of 0.8%, which, however, makes it reasonable to assume that further homozygosity in the mice may be incompatible with survival. It is obvious that producing more mice in this study would have increased the chance of finding individual mice with a high similarity between them. However, the low level of heterozygosity and the total lack of response to selective breeding indicate that further inbreeding does not seem to be a feasible method for increasing gut microbiota similarity in mice. The reason why offspring of different mothers being sisters have a higher gut microbiota similarity than if the mothers are not sisters ⁹ may be due to factors such as imprinting ¹⁷ or maternal differences in intrauterine conditions. In general, only one of the two X-chromosomes is active. ¹⁸ Sisters have inherited the maternal X-chromosome, which is known to be the most important one in relation to placenta function ¹⁹ from the same mother. An alternative for producing mice with a high gut microbiota similarity may be to inoculate these when still germfree. However, this is probably only possible within a short time window around weaning. ¹⁵