Pawedness reflects the preferential use and/or an increased capacity to perform tasks more efficiently with a specific paw. It corresponds, in general terms, to animals’ handedness. Pawedness/handedness is thought to be associated with brain asymmetries, present both at morphological, cellular and molecular levels (see for reviews Galaburda et al., 1978; Toga and Thompson, 2003; Sun and Walsh, 2006; Rogers, 2009, 2014; Hugdahl, 2011). Regarding morphology most studies have so far excluded any association (Good et al., 2001; Narr et al., 2007; Guadalupe et al., 2014, 2016; Ocklenburg et al., 2016); however at cellular and molecular levels, contralateral parietal spine density has been linked to skilled reaching (Ambeskovic et al., 2017) and dopaminergic system lateralization has been shown to be associated with hand/paw preference in humans (de la Fuente-Fernández et al., 2000) as well as in rodents (Uguru-Okorie and Arbuthnott, 1981; Schwarting et al., 1987; Barnéoud et al., 1990; Cabib et al., 1995; Nielsen et al., 1997; Budilin et al., 2008). Additionally, peripheral human (Lengen et al., 2009) and central rodent (Neveu, 1990; Shen et al., 2005) data have shown differences in the immunological system, while additional monoamines (norepinephrine; Barnéoud et al., 1990) and enzymes (angiotensinases; Wu et al., 2010) were also associated with rodent pawedness. It has therefore been hypothesized that differences in this trait might be associated with other behavioral outcomes particularly cognition. Indeed, a small advantage of right-handed people in spatial ability has been reported (Somers et al., 2015) and pawedness/memory associations have been found in monkeys (Hopkins and Washburn, 1994) and mice (Wu et al., 2010). Furthermore, Prichard et al. (2013) reported that cognitive data is associated with the degree of handedness (and not its direction) as inconsistent handedness seems to be related with better episodic memory and improved belief updating/cognitive flexibility.

While handedness assessment in humans is simple, determination of associated behavioral and molecular differences poses several challenges: (i) the distribution of left- and right-handers in the population is uneven (Teng et al., 1976; Guadalupe et al., 2016) imposing the creation of specific left-enriched cohorts; (ii) social context may alter behavior (Teng et al., 1976), therefore increasing the percentage of strong right-handers and weak left-handers; and (iii) assessment of central molecular correlates is limited. The utilization of animal models became therefore very useful in this regard.

The Collins’ (Collins, 1968) and the lateral paw preference tests (LPP; Waters and Denenberg, 1991) are amongst the most used tests to assess pawedness in rodents. The number of times a paw is used to retrieve food from an elevated tube and the amount of food retrieved from two lowered hoppers placed side-by-side are employed, respectively, as behavioral readouts. Despite the construct similarities, different results have been obtained between these two tests (Waters and Denenberg, 1991, 1994; Rogers and Bulman-Fleming, 1998) not only at the population level but, more importantly, at the individual level. Furthermore, both tests rely in the exclusive right/left paw use disregarding paw movement precision and possible intermediary strategies implicating the simultaneous use of both paws. Additionally, as measured by these methods, preference in rodents appears to a certain extant to be training/learning dependent (Collins, 1988; Stashkevich and Kulikov, 2001; Tang and Verstynen, 2002; Ribeiro et al., 2011).

To surpass these limitations, we have designed and validated an alternative test requiring minimal equipment for a fast determination of pawedness degree—the pawdeness trait test (PaTRaT). In this test, a circular grid separates the animal from a receptacle containing sugared items. Contrary to previous tests, the PaTRaT allows the simultaneous use of the two paws for reward handling reducing potential selection biases and making it more ethologically relevant. Also, the large size of the reward compared to the grid slits, imposes higher movement complexity for successful retrieval. Finally, the PaTRaT uses a classification system for dexterity degree that goes beyond the simple quantification of left-/right-paw retrievals. Additionally, we assessed potential associations between pawedness and behavioral outcomes, namely with impulsivity, memory and fine motor skills.

PaTRaT is a grading system to evaluate pawedness in rats. It is of simple implementation, requiring minimal equipment. Also, it has high inter-rater reliability and temporal stability of the outcomes. The PaTRaT displays several advantages over the most used tests—the Collins’ (Collins, 1968) and LPP (Waters and Denenberg, 1991) tests, namely: (i) the reward can be readily seen, smelled and touched by the animals, increasing motivation. Indeed, we observed in preliminary assays that animals were more prone to perform the test when a greater amount of reward (≈50 vs. ≈20 cheerios^®) was available (data not shown); (ii) the apparatus imposes no constraints to the simultaneous use of both paws for reward manipulation, decreasing potential selection biases (Ribeiro et al., 2011) and providing a more ethological setting; (iii) the large reward size relative to the mesh grid imposes higher dexterity, i.e., a sequence of relatively complex movements is required for a successful retrieval of the reward; (iv) assessment is based on four independent sessions (10 trials each) avoiding potential confounders as limb alternation due to tiredness and satiation (Bulman-Fleming et al., 1997); (v) the grading system is of simple application and has a high inter-rater agreement even among inexperienced observers; and finally (vi) PaTRaT side preference index, contrary to the simple quantification of left/right paw retrievals, accounts for intermediate pawedness and is therefore more akin to human handedness assessment (Prichard et al., 2013). Additionally, two sessions were enough to achieve a sustained performance in 80% of the animals. In our study, the prevalence of right- and left-pawed rats was similar (54% vs. 46%, respectively), which is in accordance with a previous report (Stashkevich and Kulikov, 2001) but in opposition with others that report a rightward bias up to ≈82.4% (Pençe, 2002; Elalmis et al., 2003; Güven et al., 2003; Wu et al., 2010). Scores were stable across the three evaluations even when other manipulations/behavioral paradigms were interposed; potential differences in fine motricity (or even in general motor performance) were experimentally excluded. Importantly, we evaluated the PaTRaT in male Sprague-Dawley rats but strain, sex and species differences have been described in the context of paw preference protocols (see for example Betancur et al., 1991).

It has been argued that a less marked lateralization in rodents (or other animals) in comparison with humans results from the fact that pawedness assessments rely on the evaluation of simple movements (e.g., grabbing food) instead of fine movements as in humans (see for review on nonhuman primates, Hopkins, 2013). Thus, the increased movement complexity of the PaTRaT assay may explain the absence of ambidextrous animals, normally reported as being 7.4%–23% (Stashkevich and Kulikov, 2001; Pençe, 2002; Elalmis et al., 2003; Güven et al., 2003; Wu et al., 2010), as it allows a better separation of left- and right-pawed animals (i.e., average scores close to the limits of the scale). In fact, decreases in the PaTRaT absolute scores were mostly related with poorer performances of the preferred paw and not with retrievals with the non-preferred paw (practically absent).

Pawedness in rats has been associated with central asymmetries in monoamines, notably dopamine and other molecular players (Barnéoud et al., 1990; Cabib et al., 1995; Budilin et al., 2008). These have been hypothesized to underlie associations between pawedness direction (and/or magnitude) and performance in several behavioral domains (Wu et al., 2010)—see also Prichard et al. (2013) for a comprehensive review in human studies. Specifically regarding impulsive behavior, several studies have demonstrated an influence of the dopamine levels (see for review, Dalley and Roiser, 2012; D’Amour-Horvat and Leyton, 2014). We have nevertheless observed no differences between left- and right-biased animals on impulsive behavior both on the learning protocol or in the VDS test. Additionally, no differences were observed in long-term and working spatial reference memories and reversal learning.

In conclusion, the PaTRaT is a simple, inexpensive and reliable test for assessment of pawedness degree and direction in rats. It relies on a grading system for hind paw use and has high inter-rater agreement (even for inexperienced observers). At the population level, we observed a nearly equal distribution of left- and right-biased rats and individual preferences were stable across sessions.

AMC, ME and HL-A designed the experiment; AMC, ME and MRG performed the experiments, AMC, ME, SPN and SB rated the animals; ME and HL-A performed the statistical analysis; AMC, ME, NS, AA and HL-A analyzed and interpreted the data; AMC, ME and HL-A wrote the article’s initial version. All authors revised and approved the final version of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.