動物実験：保健・生物医学研究のデザイン

The Challenge

Research using animals has been vital to Western science and medicine since its inception. Until the 1960s, however, the sex of animals used in research was rarely reported except in experiments related to reproduction. Even today, the sex of animal subjects is “omitted in 22–42% of articles in neuroscience, physiology, and interdisciplinary biology journals” (Beery et al., 2011).

Analysis of animal studies in which sex is reported shows that females are underrepresented in most subfields except reproductive biology and immunology—see chart below.

The Underrepresentation of Female Animals

Researchers may perform single-sex animal studies to reduce the cost of experiments or in hopes of lowering the variance of results (McCarthy et al., 2002). Single-sex experiments are the only option in studying sex-specific phenomena (ovarian cancer or prostate cancer, for instance) and can also be beneficial if one sex has been understudied or if there is strong evidence that sex does not influence outcome. The majority of single-sex animal experiments, however, do not fall into these categories. Female animals are underrepresented in studies of conditions that affect both sexes and are underrepresented in research where evidence suggests that sex influences outcome.

Researchers may avoid using female animals because hormone levels, which fluctuate throughout the estrous cycle, can interact with experimental outcomes (Becker et al., 2005; Wizemann et al., 2001). More rarely, researchers may avoid using male animals because, in some species and strains, inter-male aggression makes caging difficult (Gatewood et al., 2006). Female rodents may be preferred in toxicology studies because of their greater sensitivity to some toxins (European Commission, 2008).

Gendered Innovation 1: Studying Sex leads to New Treatments for Traumatic Brain Injury

Traumatic brain injury (TBI) is more common in men than in women both in Europe (where the leading cause is motor vehicle collisions) and in the United States (where the leading cause is firearm injuries; Tagliaferri et al., 2006; Wagner et al., 2000; Roof et al., 2000). New studies of TBI that include female animals have allowed sophisticated sex analysis and produced innovations in treatment for TBI patients (see Method).

Researchers used evidence from animal models of TBI—obtained through analyzing sex and sampling female animals in different hormonal states—to devise an experimental treatment for humans. In double-blind clinical studies, patients who received progesterone shortly after emergency treatment for TBI had lower mortality and showed better recovery of neurological function than control patients with similar injury, and progesterone was well-tolerated (Xiao et al., 2008; Wright et al., 2007). More research is needed to:

• Evaluate risks and benefits of progesterone treatment according to patient characteristics (such as sex and age) and injury characteristics.


• Elucidate the mechanisms by which progesterone protects against brain damage in TBI.

Gendered Innovation 2: Sampling with Attention to Estrous Cycles and Menopause Advances Basic Knowledge of the Immune System

By including female mice in experiments, scientists discovered that sex hormones are important to immune system function. When female mice were exposed to antigens and then sampled during diestrus or estrus, their immune responses in the spleen were similar to those seen in males. But when female mice were sampled during proestrus or metestrus, their antibody counts were more than triple that of males (Krzych et al., 1978). By correlating these differences with progesterone and estrogen concentrations (which also vary throughout the estrous cycle), it was possible to uncover the influence of sex hormones on immune function (Bergman et al., 1992).

Animal models of menopause—which are still in development (see Next Steps below)—have shown that immunological changes accompany this hormonal transition. When mice are ovariectomized and undergo "acute menopause," they exhibit "reduced lymphocyte chemotaxis, mitogen-induced T cell proliferation responses, and [Interleukin-2] production" (Marriott et al., 2006).

An understanding of hormones and immune function is relevant to treating numerous diseases, including autoimmune diseases and HIV infection. For example, animal models have been used to investigate why HIV viral load tends to increase more rapidly in men than in women (Meier et al., 2009).

Gendered Innovation 3: Strengthening Environmental Health Standards

In addition to their use in basic research and preclinical testing, animal models are integral to environmental monitoring and evaluating the toxicity of chemicals. The European Commission's Institute for Health and Consumer Protection (IHCP) and the U.S. National Toxicology Program (NTP) have analyzed reference models to improve environmental standards (see Method).

Standards and Reference Models in the EU

The European Commission’s IHCP has strong requirements for sex analysis and sampling (European Commission, 2008):

1. Inclusion of Female and Male Animals. For example, in inhalation toxicity testing, researchers are instructed to use equal numbers of females and males at each concentration level. In other tests, the sex more sensitive to a particular toxin, generally females, is preferred.

2. Reporting the Sex of Study Subjects. Regardless of whether an experiment is single-sex or mixed-sex, the IHCP requires reporting “number, age, and sex of animals”.

3. Sampling Pregnant Females to Detect Developmental Toxicity. This protocol allows researchers to gather “information concerning the effects of prenatal exposure on the pregnant test animal and on the developing organism in utero”.

Standards and Reference Models in the U.S.

The U.S. National Toxicology Program (NTP) requires sex analysis in all routine animal toxicity studies (NTP, 2006), which must:

1. Report the Sex of Study Subjects. NTP states that “data are to be tabulated and organized by species, sex, and treatment group.”

2. Analyze Results by Sex and Report Null Findings. All analyses performed by the NTP note the presence, absence, and statistical significance of sex differences.

3. Analyze Factors Intersecting with Sex. All analyses are to be controlled for weight so that weight differences are not misreported as sex differences or vice versa.

Conclusions

Gendered innovations:

1. Physiology and Pathophysiology: Including female animals in experimental studies led to new knowledge about traumatic brain injury, resulting in new therapeutics. Sampling with regard to sex and hormonal state has also produced knowledge about hormonal regulation of the immune system, relevant to the treatment of autoimmune diseases and infections. This information is now being applied to development of treatment dosing for vaccines (Klein et al., 2010; World Health Organization, 2010).

2. Regulatory Policy: Analyzing sex has become a critical part of both EU and U.S. efforts to understand, prevent, and control environmental pollution.

Next Steps

A. Future research needs include:

1. Analyzing Sex at the Tissue and Cellular Level. Sex analysis in basic research has occurred primarily in animal studies and has centered on hormonally mediated sex differences. Sex is rarely analyzed or even reported in studies involving cultured cells or extracted tissues. A study of articles in high-impact, peer-reviewed cardiovascular disease journals showed that only 20-28% of articles describing research on new cell lines stated the sex of cells used. Of the minority of studies that did report sex, 69% used male cells only (Taylor et al., 2011). This disparity is of concern because emerging research suggests that studying cellular sex is important in developing stem cell therapies (see Case Study: Stem Cells).

2. Analyzing Factors Intersecting with Sex to Avoid Overemphasizing Sex Differences. Not all observed differences between female and male animals, cells, or tissues—or between women and men—are due to biological sex. Analyzing factors intersecting with sex and gender is critical to avoid overemphasizing sex differences. Important factors include diet, hormone levels, and species. Maternal interactions shortly after birth contribute to sex differences in behavior: Mother rats interact differently with female and male pups, producing developmental differences (Moore, 1992).

3. Developing Animal Models of Menopause. High-quality, validated animal models of menopause are needed. Although primates undergo menopause-like processes, the challenges of using primates and the scarcity of older female animals limit research. Menopause can be induced surgically in experimental animals through ovariectomy, which models bilateral oophorectomy in women but may not be comparable to natural human menopause (Bellino et al., 2003). Strategies for modeling human menopause in rodents include treating mice or rats with drugs that induce premature ovarian failure and using transgenic mice (such as strain Foxo3a-/-) that show accelerated ovarian senescence (Wu et al., 2005).

4. Studying Gender in Animal Research. Placing female and male animals in different physical and social environments can have marked effects on behavior and experimental outcome, and gender analysis is needed to ensure that housing systems and handling do not create systematic bias (Holdcroft, 2007). In particular, if researchers expect a particular sex difference, they may handle or house female and male animals differently and in such a manner as to produce that sex difference, or they may choose a specific behavioral test likely to produce that difference (Birke, 2011). Housing and handling can determine animal stress levels, which alter both behavioral and biochemical profiles (Beck et al., 2002).

B. Policy next steps include:

1. Requiring Researchers to Report the Sex of Subjects. Granting agencies and journal editors can make sex reporting a requirement if research is to be funded or findings published. Reporting the sex of model organisms prevents inappropriate generalizations, facilitates meta-analysis, and can show where animals of one sex have been overlooked. Major bioscience funders, including the U.K.’s Medical Research Council (MRC), now require that researchers report animal “species, strain, sex, developmental stage […] and weight.” Major journals, including Nature and the Public Library of Science publications, have instituted the same requirements (National Centre for the Replacement, Refinement, and Reduction of Animals in Research, 2008; Kilkenny et al., 2010).

2. Requiring Two-Sex Studies and Sex Analysis. Government agencies can require that, where appropriate, publically funded studies include animals of both sexes and be designed with “adequate sample sizes” for each sex. Public and private funders alike can “treat inclusion of female animals as a matter of scientific merit that affects funding” (Beery et al., 2011).

3. Standardizing the Use of “Sex” and “Gender” in Relation to Animal Research. Currently, the terms “sex” and “gender” are used interchangeably in much animal research, complicating literature searches and meta-analysis. Sex and gender are not interchangeable. Standardizing usage, and using “sex” to refer to the biological trait of femaleness or maleness, would remedy this problem (Marts, 2004).