Abstract
Changes made in the 8th edition of the Guide for the Care and Use of Laboratory Animals included new recommendations for the amount of space for breeding female mice. Adopting the new recommendations required, in essence, the elimination of trio breeding practices for all institutions. Both public opinion and published data did not readily support the new recommendations. In response, the National Jewish Health Institutional Animal Care and Use Committee established a program to directly compare the effects of breeding format on mouse pup survival and growth. Our study showed an overall parity between trio and pairwise breeding formats on the survival and growth of the litters, suggesting that the housing recommendations for breeding female mice as stated in the current Guide for the Care and Use of Laboratory Animals should be reconsidered.
In 2011, the 8th edition of the Guide for the Care and Use of Laboratory Animals (the Guide) (1) was published under some perceived controversy. Although many of the changes from the previous version of the Guide were already being used by the animal research community, new recommended minimum mouse housing space requirements represented a significant and costly change from current practices. Specifically, the new Guide increased its recommended amount of space for a breeding female plus her litter to 51 in2. Standard caging for housing mice is roughly 75 in2, and although the recommendations for cage space are not strictly additive, the new recommendations created a concern about cage space when two females and two litters were housed in one cage (102 in2 > 75 in2). Prior to publication of the 8th edition of the Guide, many institutions allowed trio breeding of mice in standard cages, allowing two females and their litters to be housed together until the pups reached weaning age. Researchers from institutions across the country wrote in during the public comment period prior to publication of the new version of the Guide, expressing the opinion that the new space recommendations did not reflect their decades-long experience with mouse breeding; in fact, conventional experience suggested that housing multiple females with litters per cage benefitted the pups by reducing nervous dam-related adverse events such as abandonment or cannibalism. Further complicating matters, the literature was at the time, and remains, contradictory with regard to optimal housing densities for the health and welfare of the animals, the conclusion varying largely based on the performance criteria that were used in establishing the health and welfare of the animals (2).
To comply with the recommendations in the new Guide, all institutions were required either to change their existing trio breeding policies to reflect the new recommendations or to demonstrate that trio breeding (in standard 75 in2 caging) did not adversely affect animal welfare compared with the recommended housing density. At least initially, the Office of Laboratory Animal Welfare (OLAW) did not consider the use of literature references as sufficient scientific justification for a departure from the new recommendations.
In response to the position taken by the National Institutes of Health/OLAW, the National Jewish Health (NJH) Institutional Animal Care and Use Committee (IACUC) established a program to directly compare the impact of pairwise and trio breeding on mouse pup survival and growth, using these parameters as indicators of animal welfare. The broad conclusion of our study analyzing a total of 472 litters is that trio breeding had a positive impact on the survival and growth of the litters.
Materials and Methods
All animals at NJH are maintained in an American Association of Laboratory Animal Care, internationally accredited facility that holds Public Health Service assurance. All procedures and practices for animal use were approved by the NJH IACUC. Animals were housed in standard 75-in2 static cages. The cages and bedding (Harlan, Aspen Sani-Chip bedding) were autoclaved prior to use. Mice were given acidified water (pH 2.3–2.8) and irradiated diet (Harlan 2918). All husbandry practices aside from housing density were in accordance with the Guide. Cages were changed once or twice per week, depending on need. All cages were supplied with an autoclaved enrichment hut/igloo and half a Nestlet.
In accordance with experimental guidelines established by the IACUC, we initiated a breeding test program comparing pairwise (one male plus one female) and trio (one male plus two females) breeding formats on a broad spectrum of strains. For the trio breeding format, upon the birth of litters to both females, the male was removed from the cage until after the weaning of at least one of the litters. Importantly, our program assessed performance standards deemed by the IACUC to be consistent with recommendations in the updated Guide. These criteria were litter size, litter survival, and litter weight (average pup weight).
Rather than attempt to collect data on all strains of mice at the institution, the IACUC divided the available strains into broad categories that best represented the expected health effects of the underlying genotype. For example, the category “autoimmune” applied to mice that were prone to autoimmunity (e.g., NOD) and that might be expected to present with health complications related to autoimmunity. Similarly, the category “lymphocyte transgenic” applied to strains with T or B cell repertoires limited by the transgenic expression of a particular TCR or BCR. Representative strains from each of these categories were selected and bred in pairwise and trio combinations. A list of all strains used in the study and analysis of the breeding results are provided in Table I. Performance criteria were analyzed using an approximate (999,999 replications) two-sample permutation method (3, 4).
Table I.
List of strains used in breeding comparisons
| Formal Strain Name | Abbreviated Name | Phenotype Description/Category |
|---|---|---|
| NOD/ShiLtJ | NOD | Autoimmune |
| B6.NZB-(D1Mit47-D1Mit209)/BkotJ | B6.Nba2 | Autoimmune |
| NOD/ShiLt-Tg(RipTAg)1Lt/J × NOD)BC | RipTag | Autoimmune |
| B6.Cg-Sle1NZM2410/Aeg Yaa/DcrJ | B6.SLE | Autoimmune |
| NOD.CB17-Prkdcscid/J | Nod Scid | Immune deficient |
| B6.129-Ifnar1tm1Agt | IFNαR−/− | Immune deficient |
| B6.129-Ifngtm1Agt | IFNγ−/− | Immune deficient |
| B6.129S7-Rag1tm1Mom/J | B6 Rag1 | Immune deficient |
| inpp5dfl/flxhCD20tamcrexB6.IgHArs/A1 | Ars/A1 SHIP-1 cKO | Lymphocyte transgenic |
| A.B6- Igkctm1CgnTg(IgkAb36-71)1Wys | A/J ktg k−/− | Lymphocyte transgenic |
| B6.CgTg (TcraTcr6)425 | 425Tg | Lymphocyte transgenic |
| C57BL/6-Foxp3tm1Flv/J B6.CgTg (Tcrb)Y48A | Y48A Foxp3 GFP | Lymphocyte transgenic |
| C57BL/6J-Tg(Itgax-cre,-EGFP)4097Ach/J | CD11c-Cre GFP | WT |
| C57BL/6-Tg(UBC-GFP)30Scha/J | UBI-GFP | WT |
| C57BL/6J | B6 | WT |
| C57BL/6J × A/J | B6AF1 | WT |
| BALB/cJ | BalbC | WT |
| B6.Cg-Tg(Prdm1-EYFP)1Mnz/J | Blimp-YFP | WT |
| FVB/N-Tg(Tagln-rtTA)E1Jwst/J | Sm22-rtTa | WT |
| B6.CgGt(ROSA)26Sortm1(rtTA,EGFP)Nagy/J | Rosa26-rtTA | WT |
| B6;129S4-Gt(ROSA)26Sortm1Sor/J | Rosa26-LSL-LacZ | WT |
| Tg(Fos-lacZ)34Efu/J | TOPGAL | WT |
| Tg(Fgf10LacZ)Mb | Fgf10LacZ | WT |
| B6.129-Ctnnb1tm2Kem/KnwJ | β-cat(ex3)/(ex3) | WT |
| B6-Ilktm2/Rf | Ilkf/f | WT |
| B6;129S-Snai1tm2Grid/J | Snai1f/f | WT |
| Stock-Fgf10tm2/Sm | Fgf10f/f | WT |
| B6.CgGt(ROSA)26Sortm1(rtTA,EGFP)Nagy/J | Rosa26-LSL-rtTA | WT |
| B6.129 × 1-Twist2tm1.1(cre)Dor/J | Dermo1-Cre | WT |
| B6.Cg-Tg(Myh11-cre,-EGFP)2Mik/J | smMHC-Cre-IRES-GFP | WT |
| B6.129-Nkx2.1tm1(Cre)Bs | Nkx2.1-Cre | WT |
| Wt1tm2(cre/ERT2)Wtp/J | Wt1-CreERT2 | WT |
| Wt1tm1(EGFP/cre)Wtp/J | Wt1-Cre | WT |
| B6.Cg-Shhtm1(EGFP/cre)Cjt/J | Shh-Cre | WT |
| B6-Scgb1a1tm1(Cre)Fdm | Scgb1a1-Cre | WT |
| Ascl1tm1.1(Cre/ERT2)Jejo/J | Ascl1-CreERT2 | WT |
| B6-α-SMAtm3(Cre/ERT2)Pch | α-SMA-CreERT2 | WT |
| Gli1tm3(cre/ERT2)Alj/J | Gli1-CreERT2 | WT |
| B6.Cg-Tg(Col1a2-cre/ERT)7Cpd/J | Col1a2-CreERT | WT |
| B6;129S-Sox2tm1(cre/ERT2)Hoch/J | Sox2-CreERT2 | WT |
| B6.129 × 1-Twist2tm1.1(cre)Dor/J | Scgb1a1-CreER | WT |
| Tg(Tet-Dkk1)Sdl | Tet-Dkk1 | WT |
| Tg(Tet-Fgf10)Jw | Tet-Fgf10 | WT |
| Tg(Tet-sFgfr2b)Jw | Tet-sFgfr2b | WT |
| FVB/N-Tg(tetO-MYC)36aBop/J | Tet-cMyc | WT |
| Tet-Cre B6.Cg-Tg(tetO-cre)1Jaw/J | Tet-Cre | WT |
List of strains used in breeding study. Assignment of “category” was determined as described in text. WT, wild-type.
Results
At the end of the data collection period (~6 mo), data on a total of 472 litters were analyzed to determine whether any of the performance criteria varied in a statistically reliable fashion relative to breeding format. The data were assessed in three cohorts. First, we compared the two breeding formats across all strains in the data set (Table I). Although there were no statistically significant differences between the two breeding formats with regard to average pup weight, trio breeding actually provided a statistically significant increase in both litter size and litter survival (Fig. 1). Second, we examined the data with strains grouped into the specific categories to which they were assigned. Analyzed in this fashion, trio and pairwise breeding strategies were statistically indistinguishable from one another in all performance criteria (Table II). Lastly, we compared trio versus pairwise outcomes within a variety of single strains. Again, no statistically significant differences were found between trio and pairwise breeding formats at this level (Table III). As expected, note that certain strains and categories had growth profiles commonly associated with their specific genetic alterations. For example, strains of mice in the “autoimmune” category typically had lower average pup weights at 21 d when compared with strains that were phenotypically wild-type. However, none of these within-strain criteria varied relative to breeding format.
FIGURE 1.
For 472 litters of all strains, litter size, percentage survival, and average pup weight in grams at 21 d of age in duos (n = 197) and trios (n = 275) are shown. Litter size (p < 0.001) and percentage survival (p < 0.001) were greater in trios. Average pup weight at 21 d was similar in duos and trios (p = 0.24). Medians are depicted in red. The blue curves represent the smoothed distributions of values.
Table II.
Breeding analysis for strain categories
| Litter Size | % Survival | Average Pup Weight (g) |
||||
|---|---|---|---|---|---|---|
| Category | Duo | Trio | Duo | Trio | Duo | Trio |
| 1. Autoimmune (25D/66T) | 8 | 7 | 89 | 89 | 8.8 | 9 |
| 1. Lymphocyte transgenic (14D/14T) | 3 | 6 | 100 | 100 | 9.4 | 10.7 |
| 2. WT (19D/31T) | 6 | 7 | 100 | 100 | 11.2 | 10.8 |
Breeding statistics for strains grouped into specific categories as described in Table I. All entries are median values. Average pup weight shown is in grams at 21 d of age. Number of litters in each breeding format used in the analysis are shown in parentheses and formatted as no. of litters in duo (D) breeding/no. of litters in trio (T) breeding. For all categories, litter size, percentage survival, and average pup weight are not statistically different in trio compared with duo. WT, wild-type.
Table III.
Single strain breeding analysis
| Litter Size | % Survival | Average Pup Weight (g) |
||||
|---|---|---|---|---|---|---|
| Strain | Duo | Trio | Duo | Trio | Duo | Trio |
| 1. B6AF1 (8D/10T) | 6 | 7 | 100 | 100 | 11 | 9.9 |
| 2. CC10Cre … ILK (2D/7T) | 8 | 6 | 50 | 89 | 14.9 | 11.5 |
| 3. GKO (6D/6T) | 7.5 | 9.5 | 94 | 100 | ND | ND |
| 4. NOD (9D/15T) | 9 | 8 | 100 | 100 | 10 | 10 |
| 5. Rosa ILK (13D/16T) | 4 | 6.5 | 100 | 100 | 10.7 | 11.8 |
| 6. sm22rtta … lacz (5D/5T) | 6 | 6 | 83 | 100 | 11.7 | 11 |
| 7. B6.Nba2 (10D/24T) | 6 | 7 | 71 | 80 | 7.5 | 8 |
| 8. Blimp (8D/18T) | 7.5 | 5 | 59 | 82 | 8.4 | 8.5 |
| 9. LSC (4D/15T) | 5.5 | 5 | 44 | 80 | 7.2 | 8 |
Breeding statistics for individual strains. All entries are median values. Average pup weight shown is in grams at 21 d of age. Number of litters in each breeding format used in the analysis are shown in parentheses and formatted as no. litters in duo (D) format/no. of litters in trio (T) format. For all strains, litter size, percentage survival, and average pup weight are not statistically different in trio compared with duo. For the GKO strain, pup weight at day 21 was not determined (ND).
Although our data collection did not capture results on all possible backgrounds, phenotypes, and strains bred by this institution, it did include strains from at least four inbred backgrounds (B6, BalbC, B6xA/JF1, and NOD) and a broad spectrum of genetic alterations. No statistically significant differences in the performance criteria tracked with genetic background or genetic alteration when comparing pairwise and trio breeding formats. Importantly, incorporating more strains with an increasing diversity of genetic backgrounds into the analysis resulted in a statistical increase rather than a reduction in the advantage of trio over pairwise breeding (Fig. 1).
Discussion
The new edition of the Guide contains changes from the previous version, commonly expressed in the form of “must” and “should” statements. Broadly defined, “must” statements reflect animal care provisions that are mandatory whereas “should” statements are recommendations about animal care. It is generally assumed that institutions comply with must or mandated changes to animal care, whereas should or recommended changes are followed unless the institution has sufficient scientific justification for modifying its response to the recommended change. Many of the new Guide changes were uneventfully received by the research community, in part because many of the changes were already standard practices in research institutions (e.g., animal enrichment devices in rodent cages). An exception was the change to the recommended cage space for breeding female mice. Many in the research community objected to this specific recommendation, largely on the grounds that requiring such specific rodent housing densities was unsupportable given that much of the available literature at the time seemed to contradict any rationale for the recommendation (5–9). Note that these objections have been fortified by data published after implementation of the Guide as well (10–12).
In its communications to the public during the initial implementation of the new edition of the Guide, OLAW reiterated the assertion that the new housing recommendations were not “must” statements and were subject to revision by any institution provided that sufficient scientific justification was available to support the revisions. As usual for animal welfare considerations, cost or convenience could not be used a basis for justification, although estimates of the cost of conforming to the new standards were significant. Furthermore, OLAW stated at the time that it did not consider the use of literature references as sufficient scientific justification, presumably due to their conflicting nature (2).
For these reasons, the NJH IACUC initiated and administered a 6-mo breeding program to determine whether any scientific justification could be identified for requiring pairwise breeding over trio breeding in standard 75-in2 caging. Our overall conclusion from our data is that there is parity between trio and pairwise breeding formats for mice across a broad spectrum of strains and genetic alterations. As they apply to the Guide recommendations, these data indicate that the recommended space for a breeding female should not be treated as additive, at least for two females per 75-in2 cage. Somewhat surprising was the statistical advantage of trio over pairwise breeding schemes as the size of the data set increased. The parameters of trio relative to pairwise breeding that might be responsible for this advantage were not identified in our study, but they warrant further investigation.
The Animal Welfare Act requires that any principal investigator performing animal research must “… [provide] written assurance that the activities do not unnecessarily duplicate previous experiments,” a requirement also stipulated in the Guide (Ref. 1, p. 25). In light of this mandate, it was somewhat surprising to the research community that OLAW has not supported the use of literature references alone as justification for deviations from the new housing recommendation for a breeding female. Indirectly, this requires the generation of new data within each institution, the acquisition of which is in apparent contradiction to the mandate against unnecessary duplication of research, especially given the available literature on the topic (5–12). Although performance criteria may vary to some degree between different institutions due to a variety of factors (e.g., environment, husbandry practices, altitude, humidity), the likely range of variation contraindicates engaging in individual studies at each institution on both practical and ethical grounds. The NJH IACUC communicated its findings to OLAW along with its conclusions and the intent to broadly approve the use of trio breeding policies within the NJH animal facility. In a return communication, OLAW concurred with the action taken by the IACUC and stated that the NJH process supported scientific objectives and the health and welfare of the animals. Given these statements, and the requirement of all animal researchers to avoid unnecessary duplication of published literature, we think that the data presented in the present study should serve more broadly as a useful reference to all investigators engaged in animal research for the purposes of breeding colony management.
Abbreviations used in this article
- Guide
Guide for the Care and Use of Laboratory Animals
- IACUC
Institutional Animal Care and Use Committee
- NJH
National Jewish Health
- OLAW
Office of Laboratory Animal Welfare
Footnotes
Disclosures
The authors have no financial conflicts of interest.
References
- 1.National Research Council (U.S.) Committee for the Update of the Guide for the Care and Use of Laboratory Animals. Guide for the Care and Use of Laboratory Animals. 8th Ed. Washington, DC: National Academies Press; 2011. [Google Scholar]
- 2.Whittaker AL, Howarth GS, Hickman DL. Effects of space allocation and housing density on measures of wellbeing in laboratory mice: a review. Lab. Anim. 2012;46:3–13. doi: 10.1258/la.2011.011049. [DOI] [PubMed] [Google Scholar]
- 3.Curran-Everett D. Explorations in statistics: permutation methods. Adv. Physiol. Educ. 2012;36:181–187. doi: 10.1152/advan.00072.2012. [DOI] [PubMed] [Google Scholar]
- 4.Good P. Permutation, Parametric, and Bootstrap Tests of Hypotheses. New York: Springer; 2005. [Google Scholar]
- 5.Davidson LP, Chedester AL, Cole MN. Effects of cage density on behavior in young adult mice. Comp. Med. 2007;57:355–359. [PubMed] [Google Scholar]
- 6.Laber K, Veatch LM, Lopez MF, Mulligan JK, Lathers DM. Effects of housing density on weight gain, immune function, behavior, and plasma corticosterone concentrations in BALB/c and C57BL/6 mice. J. Am. Assoc. Lab. Anim. Sci. 2008;47:16–23. [PMC free article] [PubMed] [Google Scholar]
- 7.Nicholson A, Malcolm RD, Russ PL, Cough K, Touma C, Palme R, Wiles MV. The response of C57BL/6J and BALB/cJ mice to increased housing density. J. Am. Assoc. Lab. Anim. Sci. 2009;48:740–753. [PMC free article] [PubMed] [Google Scholar]
- 8.Smith AL, Mabus SL, Muir C, Woo Y. Effects of housing density and cage floor space on three strains of young adult inbred mice. Comp. Med. 2005;55:368–376. [PubMed] [Google Scholar]
- 9.Smith AL, Mabus SL, Stockwell JD, Muir C. Effects of housing density and cage floor space on C57BL/6J mice. Comp. Med. 2004;54:656–663. [PubMed] [Google Scholar]
- 10.Horn MJ, Hudson SV, Bostrom LA, Cooper DM. Effects of cage density, sanitation frequency, and bedding type on animal wellbeing and health and cage environment in mice and rats. J. Am. Assoc. Lab. Anim. Sci. 2012;51:781–788. [PMC free article] [PubMed] [Google Scholar]
- 11.Morgan JL, Svenson KL, Lake JP, Zhang W, Stearns TM, Marion MA, Peters LL, Paigen B, Donahue LR. Effects of housing density in five inbred strains of mice. PLoS ONE. 2014;9:e90012. doi: 10.1371/journal.pone.0090012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Paigen B, Svenson KL, Von Smith R, Marion MA, Stearns T, Peters LL, Smith AL. Physiological effects of housing density on C57BL/6J mice over a 9-month period. J. Anim. Sci. 2012;90:5182–5192. doi: 10.2527/jas.2012-5417. [DOI] [PMC free article] [PubMed] [Google Scholar]