Laboratory Animals - June Issue - 248

248
inﬂammation, though it was lower than in mice with
mouse GM and despite having reduced levels of Th1
and Th2 cytokines before the model was induced.16
Hence, if the read-out of the model is of a magnitude
signiﬁcant enough to be biologically meaningful, it may
be less problematic that the immune baseline is below
that of mice with mouse GM.

Can we overcome the challenges?
Physiological, anatomical and immunological abnormalities of GF mice are well documented and
summarized elsewhere.25 Most of these features are
normalized after colonization with a complex GM.
However, irrespective of whether the GM transplant
is murine or human, a GF period in early life prior to
colonization alters immunological shaping26 and
ability to confer colonization resistance against pathogens21 later in life. Hence, to allow for correct
immunological development study mice should not be
the parent ex-GF generation, but rather the oﬀspring
generations born with the GM.21 Sex also inﬂuences
the GM proﬁle. GM from a male donor established
diﬀerently in male and female mice,27 suggesting
that alignment of donor and recipient sex may be
necessary.
Few, if any, attempts to optimize the colonization
eﬃciency in human-to-mouse GM transfer studies have
been published. Yet, several procedural steps from collection of the donor material, handling, and storage to
the transfer itself and the conditions following transfer
may aﬀect colonization eﬃciency. For practical reasons, feces is the most frequently used donor material,
but a more truthful representation of the entire gastrointestinal tract in the donor sample might also enhance
the diversity of the established microbiota. Transfer of
fresh versus frozen material works equally well,6,28 but
the exact cryoprotection protocol, anaerobic conditions, single versus repeated gavage or the use of
other administration forms may all inﬂuence transfer
rate. For instance, a model-community of 10 bacterial
strains from human feces was more successfully established in GF mice by separate monocolonizations and
subsequent co-housing compared with one single
administration of all 10 bacteria.3 Husbandry factors,
diet probably being the most important one, are likely
to impact the colonization. Genetic humanization, for
example, of epithelial colonization factors, or humanization of immune cell populations are also interesting
avenues yet to be explored. For obtaining an immunological phenotype similar to speciﬁc pathogen-free
(SPF) mice, it may be necessary to add indigenous
mouse bacteria to the human GM, as demonstrated
with segmented ﬁlamentous bacteria.29 Such hybrid
approach combines the ecological community of a

Laboratory Animals 53(3)
human GM with a more developed immune system
and may thus be a better reﬂection of the reality.

What are the alternatives?
The major limitation of mice with human GM is the
altered immunological shaping compared with mice
harboring an SPF GM. Interestingly, SPF mice are
criticized for exactly the same compared with their
wild counterparts.30-32 In our eﬀorts to exclude
unwanted microorganisms according to SPF health
standards, the GM of SPF mice seem to have lost
important immunomodulatory organisms. However, it
complicates the picture that SPF rodents harbor diﬀerent GM across animal institutions, which in turn elicit
widely diﬀerent phenotypes33,34 and make generalizations impossible. Nevertheless, the question has been
raised whether SPF mice are good models for humans
at all.30-32 We do not live in clean enclosures, and it is a
natural part of our immunological education to be
exposed to a variety of both beneﬁcial and pathogenic
microorganisms. Laboratory mice with a GM from
wild mice have signiﬁcantly improved disease resistance
due to a better poised immunological state, and provide
a provocative alternative to the too-clean SPF mice,32
and maybe also to mice with human GM. A large functional overlap between SPF GM and human GM has
been shown,35 and is an argument in favor of mouse
cohorts with a co-evolved GM (wild or traditional SPF)
that is functionally comparable to that of humans, yet
able to induce immunocompetency in the hosts.
Rats have already been mentioned as potiential superiour recipients of human GM. Pigs are often proposed as
more precise models of humans because of the highly
comparable physiology. The GM of GF piglets colonized with human GM do indeed seem to have a good
conformity with the donor.36 In the same study, the
authors did several rounds of GM administrations to
the piglets over 10 days,36 which may also have enhanced
the colonization eﬃciency compared with rodent studies
often relying on a single round of oral gavage. Another
study found that a human GM induced immune parameters in GF piglets, that is, they had more CD4þ T cells
and MHC class II expressing cells in the gut compared
with piglets colonized with pig feces.37 In this study,
Biﬁdobacterium established well in the porcine hosts
and may provide at least a partial explanation of the
results.37 Recently, a dog GM gene catalog was published and demonstrated that compared with the
mouse and pig microbiome, the dog GM was most similar to the human GM,38 implying that dogs may be a
highly translationally relevant model system for human
microbiome research. However, compared with mice,
there are other limitations related to the use of rats
and pigs and other animal models. Choice of model is
Laboratory Animals - June Issue

Table of Contents for the Digital Edition of Laboratory Animals - June Issue
