The occurrence of tarsal injuries in male mice of C57BL/6N substrains in multiple international mouse facilities

Dislocation in hindlimb tarsals are being observed at a low, but persistent frequency in group-housed adult male mice from C57BL/6N substrains. Clinical signs included a sudden onset of mild to severe unilateral or bilateral tarsal abduction, swelling, abnormal hindlimb morphology and lameness. Contraction of digits and gait abnormalities were noted in multiple cases. Radiographical and histological examination revealed caudal dislocation of the calcaneus and partial dislocation of the calcaneoquartal (calcaneus-tarsal bone IV) joint. The detection, frequency, and cause of this pathology in five large mouse production and phenotyping centres (MRC Harwell, UK; The Jackson Laboratory, USA; The Centre for Phenogenomics, Canada; German Mouse Clinic, Germany; Baylor College of Medicine, USA) are discussed.


Introduction
Inbred strains of laboratory mice are used to standardise the genetic background of mutant mouse strains to reduce data variability. Produced by >20 consecutive generations of sibling mating, the controlled homogeneity of inbred strains such as C57BL/6N is accompanied by the fixing of spontaneous mutations in inbred genomes. Many monogenic mutations have been identified in inbred mouse strains, including those causing retinal degeneration in C3H strains [1] and age-related deafness in C57BL/6 strains [2]. The characterisation of these mutations has allowed their impact on individual research programs to be assessed and alternative genetic backgrounds used if they interfered with the primary purpose of the studies. a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 Conversely, there are reports of sporadic, low-level defects in inbred lines which are likely due to oligogenic or polygenic effects and exhibit variable penetrance thus only observed or measured in a proportion of the population of an inbred colony [3]. These include complex behaviours such as aggression [4], hyperactivity [5], morphological anomalies such as sternal segment dislocation [6] and developmental defects such as hydrocephalus (https://www.jax. org/news-and-insights/2003/july/hydrocephalus-in-laboratory-mice).
Knowledge of the predisposition of mouse strains to such issues is not only essential for the care and welfare of mice but is an important consideration in phenotyping programs. It is crucial to distinguish incidental effects caused by genetic background, from outcomes arising because of an experimental paradigm (e.g. genetic mutation or physiological challenge) or a combination of background and paradigm together. The International Mouse Phenotyping Consortium (IMPC) (www.mousephenotype.org) is generating a genetically altered (GA) mouse strain carrying a null allele for each protein-coding gene in the mouse to study mammalian gene function. GA strains for this program are generated on the C57BL/6N genetic background and phenotyping is performed at an early adult time point (up to 17 weeks) and a late adult time point (after 12-18 months) for a subset of strains. Phenotyping and husbandry protocols include the regular assessment of welfare and fitness during handling and cagechanging, and motor function during phenotyping tests [7].
In this study, we report the recurrent observation of abnormal hindlimb morphology, accompanied by lameness, in group-housed male mice of C57BL/6N substrains. This report describes the nature of the injury and discusses possible aetiologies. We also provide an estimate of the frequency of occurrence from five large mouse genetics centres in four different countries across two continents and highlight potential consequences for projects where prolonged group-housing of male C57BL/6N mice is a necessity. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Ethics statement
Animal Welfare (OLAW), and meet the requirements of the Public Health Service Policy on Humane Care and Use of Laboratory Animals (assurance number D16-00475).

C57BL/6N substrains used in this study
Mice used at MRC Harwell were C57BL/6NTac, originally purchased from Taconic Biosciences, USA and subsequently bred at MRC Harwell. Mice examined at The Jackson Laboratory are sourced from an in-house maintained colony of C57BL/6NJ. Mice used at The Centre for Phenogenomics are C57BL/6NCrl, purchased from Charles River Laboratories, USA and subsequently bred at TCP. Mice at the German Mouse Clinic were C57BL/6NTac purchased from Taconic Biosciences, Germany and C57BL/6NCrl purchased from Charles River Laboratories, Germany. Mice at Baylor College of Medicine were C57BL/6NJ originally purchased from The Jackson Laboratory, USA and subsequently bred at this facility.

Mouse housing conditions
Housing conditions in each institution are listed in S1 Table. All mice were given food and water ad libitum. Adult mice were humanely sacrificed by an overdose of anaesthetic, overdose of carbon dioxide, or by cervical dislocation (according to relevant national and local protocols and guidelines).

Clinical examination
The period of data collection varied according to the schedule of the individual centres but the total number of mice presented represents the observation of all of the male mice, of the stated substrain, enrolled in the IMPC ageing program.
Routine animal care and welfare checks in all facilities involved visually inspecting the mice as part of a daily check and regular handling (typically no less than once every 14 days) during cage-changing, or during phenotyping. Mice with an abnormal gait and/or locomotor deficit accompanied by abnormal hindlimb morphology and swelling or reddening of the tarsus were euthanised for selected increased observation to ensure no further deterioration in the welfare of the mouse.

Radiography
Lateral views of the affected and contralateral tarsi were taken of representative animals under isoflurane anaesthesia by digital radiography at 26 kV for 3 s using a Faxitron MX-20 digital X-ray system or a Faxitron X-Ray Model Ultrafocus 100 (both from Faxitron X-ray Corporation, Lincolnshire, IL, USA). X-ray images were processed using the DicomWorks software (http://www.dicomworks.com/).

Histopathology
Immediately following euthanasia, tissues from selected mice were fixed in 10% neutral buffered formalin for a minimum of 24 hours. Following fixation, the hindlimbs (affected and contralateral) were stripped of soft tissues, decalcified in formic acid for 96 hours, and processed routinely for histopathologic evaluation. Subsequently, 4-5 μm thick, mid-sagittal sections were stained with haematoxylin and eosin (H&E) for evaluation. Representative images were acquired using an Olympus BX43 microscope with a Micropix Elite 5MP camera and Cytocam software v1.6. All histologic evaluations were performed by a board-certified veterinary pathologist.

Clinical examination
A proportion of male mice of C57BL/6N substrains, housed in social groups were observed to have clinical signs of abnormal hindlimb morphology, together with swelling or reddening of the tarsus and often an abnormal gait. Similar numbers of C57BL/6N females were assessed and no tarsus, hind paw, or gait abnormalities were observed. Gait abnormalities in males ranged from limping with a reduced amount of weight bearing on the affected limb to complete non-weight bearing. Grossly affected tarsi showed a loss of the abrupt right angle formed from the calcaneus and the calcaneal tendon, and there was variable soft tissue swelling sometimes accompanied with redness (Fig 1). Whilst the majority of affected mice had only one abnormal hind paw, 1/21 at MRC Harwell, 4/21 at TCP and 15/58 at JAX presented with bilateral tarsal abnormalities.
Radiography confirmed caudal dislocation of the calcaneus and new periosteal bone formation (Fig 2). In some animals, there was also calcification within the distal calcaneal tendon.

Histopathology
Histopathological examination identified caudo-dorsal dislocation of the calcaneus with concurrent partial dislocation and hyperextension of the calcaneoquartal joint. In more chronic lesions, the calcaneoquartal joint progressed to new bone formation (Fig 3). There was no difference in overall dislocation of the calcaneus in acute versus chronic lesions.

Frequency and variability of occurrence
C57BL/6N mice bred for the IMPC late adult phenotyping program were examined for tarsal injuries at five international mouse research centres. These mice included GA strains from a wide range of mutant lines examined by the IMPC as well as baseline wild type controls. Due to differences in individual institutional protocol, the ages of the cohorts vary. The frequency of occurrence was between 1.7% and 12.1% of male mice examined between the ages indicated

PLOS ONE
Tarsal injury in C57BL/6N male mice ( Table 1). The age of each individual mouse when the injury was first identified is shown in S2 Table. The frequency of affected animals reported represents those animals presenting signs of lameness and/or swelling that could be detected by visual inspection alone. Therefore, the

PLOS ONE
Tarsal injury in C57BL/6N male mice absolute frequency of occurrence will be dependent upon the subjective nature of observations by different technical staff and inspection practices in individual centres. The values presented here potentially under-estimate the total frequency of tarsal injury. Further imaging such as Xray may reveal further injuries which had subclinical or subtle presentations that were not detected by observations alone.

Husbandry and housing affected
Further observations were made and recorded which informed the aetiology of the incidence of the tarsal injury in C57BL/6N mice.
• Female mice: Equivalent numbers of females of the same strain which were part of the same program of work were also examined but no similar injury was reported.
• Males in mating cages or single-housed: No injury was observed in 584 C57BL/6NTac males in either mating cages (with one or two females) or single-housed mice examined between the ages of 16 and 64 weeks (average age 24 weeks) at MRC Harwell.

Discussion
Here we report the observation of tarsal injury in male mice with a C57BL/6N genetic background occurring at five large and geographically-dispersed mouse facilities. These injuries were observed in a number of different mutant strains and several wild type substrains indicating a predisposition for such lesions in mice of C57BL/6N ancestry. A similar deformity has been reported in STR/ort mice which have a known genetic abnormality predisposing them to chronic arthropathy and are used as a model for studying osteoarthritis [8]. In the STR/ort mice, lameness and hind paw deformity also affected predominantly male mice although the incidence rate was far higher and occurred from a younger age compared with our observations. The radiographic findings and histopathology are consistent with an injury caused by frequent high load tension from the calcaneal tendon through its insertion to the calcaneus, leading to a breakdown of the plantar ligaments supporting the calcaneoquartal joint which are weakened in this model by a known collagen abnormality [9]. There is no known underlying abnormality in the C57BL/6N substrains reported here and so it is hypothesised that the lesion is caused by application of an abnormally high load/force through the calcaneal tendon because of behavioural or husbandry practices. It should be noted that all animals in this study are fed on regular maintenance or breeding diets and not on high-fat or obesity-inducing regimes.
The type of lesion we identified was restricted to group-housed males and its occurrence became more prevalent as the mice aged. However, this may represent an increased

PLOS ONE
Tarsal injury in C57BL/6N male mice opportunity for this injury to occur over time, rather than an increased predisposition/weakness in older males. The absence of any such injury in female mice socially-housed for the same experimental purposes and for the same length of time indicates that this is a sexually dimorphic effect. As these injuries were observed in three different C57BL/6N substrains it is possible that this genetic background is predisposed to tarsal injuries. Male C57BL/6 mice are widely reported to display aggressive behaviours towards cage-mates [10]. Both threat (thrust and mounting) and aggressive behaviours (boxing, parrying, fighting) are associated with establishing and maintaining dominance hierarchies in group-housed male mice. Each of these behaviours involve rearing that requires repeated plantar flexion of the hind paw at the tarsus, initiated by high load tension from the common calcaneal tendon. Sporadic and frequent bouts of fighting have also been associated with an increased mechanical load on male tibiae in C57BL/6J mice [11], a strain related to C57BL/6N. All five centres report the observation of aggressive behaviour in maturing mice in this study. This includes a variety of injuries likely to be inflicted by fighting, such as wounds to the neck, lower back, abdomen, tail, and urogenital region. In younger cohorts of these substrains (less than 16 weeks), these types of injury are rarely seen. Increases in injuries due to aggression in aging groups may represent either an increased opportunity during long-term housing and/or a behavioural change during the aging process. Interestingly, there was no correlation between fight wounds and tarsal injuries in the mice reported. This may indicate that aggressive behaviour is not the primary cause of this injury, that the injured mice are the aggressors or that the nature of the social interaction resulting in these injuries (such as mounting or rearing in threat) do not result in biting.
It is therefore unclear whether the causal feature of the injury we observed is an inherent weakness in the tarsal joint, a consequence of a behavioural characteristic of C57BL/6N male mice, their interaction with the environment, or combinations of these factors. The variation in observed frequencies of tarsal injury between the centres may be explained by factors such as differences in housing and animal care regimes. Investigations into different husbandry protocols may provide insight into ways to reduce occurrence in the future. Euthanasia following discovery of the injury described may have substantial consequences to the study being undertaken. Disruption of an established cage-group hierarchy by removing an individual may lead to further perturbations in both the behaviour of the existing animals or to the experimental design itself with a reduction in data collected and the potential loss of statistical power. It is possible that these injuries may occur in strains of mice other than C57BL/6N and that the frequency of occurrence may be strain-dependent.
In summary, this report provides a description of an injury to group-housed male C57BL/ 6N mice observed in five different mouse centres from studies involving large numbers of animals. It is likely that a similar incidence may occur undetected, or be attributed to experimental protocols, in other facilities using these substrains. Until the exact cause of the reported injury can be identified, it may be difficult to prevent in group-housed animals. As this type of injury has never been reported in extensive monitoring of single-housed males, solitary housing may be an option. However, the issue of housing single mice for long periods of time is frequently debated, with regulatory guidelines in both Europe (https://www.coe.int/en/web/ conventions/full-list/-/conventions/treaty/123, Appendix A) and the USA [12] exonerating the importance of social housing in laboratory animals. Moreover, the loss in opportunities to examine the effects of social interaction and the potential of strong phenotypic effects of single-housed mice [13] all need to be considered carefully.
The implications of these findings will be study-dependent but have the potential to affect phenotyping results or cause an increase in attrition for ageing studies, resulting in insufficient animals completing the studies. The information reported here should be used to assist future experimental design for longitudinal studies especially those involving measurements of gait and motor skills.