Sirs

We appreciate the opportunity to emphasize that we did not intend to suggest, even indirectly, that primate research should be replaced by research on rats. We tried to make a case that the rat, if used optimally, is a relevant species for neurological research. Non-human primates will continue to be indispensable as a model for developing and evaluating the efficacy and safety of new treatments1, as well as for investigating species differences and similarities in the organization of the neural systems that control movement. We believe that there is a need for cooperation among rodent, primate and human researchers to improve the current behavioural methods. Modelling parkinsonian-like symptoms in rats has advanced substantially since the early days when only drug rotation tests were used.

We feel that there is nothing misleading in what we wrote that could potentially be interpreted as arguing that only rats should be used for basic research. However, because the letter of Lemon and Griffiths questioned how well findings from rodent studies might generalize to humans, we would like to reply briefly to some specific points they raised.

First, Lemon and Griffiths argue that whereas the limbs of primates and rodents are homologous in that they derive from a common ancestor, the function of the limbs are not. We have argued, however, that the limbs of rodents and primates have homology in function2. In both orders, the limbs are used for walking, running, climbing, grooming and manipulating objects such as food. So, we argue that for the many functions for which there is homology, the rat provides an appropriate species for modelling neurological deficits. Furthermore, we would agree that the hands of humans and of some primates are used for more complex functions than the rodent paws are, and for modelling these functions primates will clearly be more useful than rodents.

Second, Lemon and Griffiths argue that, although the neural systems of rodents and primates are homologous, they have essential structural differences and subserve different functions. To illustrate this point, they refer to the organization of the corticospinal tract, a topic that we briefly touched on in our paper3. Although we agree that the corticospinal tracts of rats and primates are not identical, they share a surprising number of anatomical similarities4. Given these similarities, it should not be surprising that there are homologies for the functions of this tract between rats and primates. Lemon and Griffiths make a strong point about the importance of the primate corticospinal tract in the control of individual digits. However, the control of digits is not its only function; this tract is recognized to be involved in many kinds of skilled movements of the extremities. There is substantial evidence to support the view that the projections of the rat corticospinal tract do not reflect only a function in the descending control of somatosensory input, but rather they also participate in the control of motor behaviour.

We refer interested readers to literature showing that stimulation of the primary motor cortex evokes limb movements, and that lesions of the corticospinal tract impair skilled forelimb movements in the rat5,6. Moreover, in both primates and rodents the cortical representation of the distal forelimb increases in size after the acquisition of a specific motor skill that requires movements of the digits and wrist7,8. So, not only does the pyramidal tract of rodents and primates show homology of function, but it also undergoes similar plastic changes in response to motor experience.

Third, in referring to the effects of motor system damage in rodents, Lemon and Griffiths argue that, compared to the effects in primates, the consequences are rather subtle, and their detection requires special tasks and sophisticated equipment. However, a central point of our argument is that if the same neurological examination that is given to a patient is applied to rodents, then surprisingly similar deficits are seen after similar damage to the nervous system. As we illustrate with the video presentations that accompanied our article3, the tasks that we have used to describe motor impairments in rats consist of describing their posture and muscle tone, how they initiate weight-shifting movements and support themselves to maintain balance, how they walk on uneven surfaces, and how they reach for food. The equipment consists largely of visual inspection, a video camera and, in extreme cases, systems for digitizing joint angles. Students of the motor system of primates and humans will recognize these tasks and equipment as being essentially the same as those that they commonly use. It is precisely because the deficits in rats can be so easily observed and scored that we argue that rats make a good model for human neurological disorders, but not to the exclusion of primates.

Last, Lemon and Griffiths suggest that non-human primate models have contributed far more toward our understanding of the pathophysiology of idiopathic Parkinson's disease than rodent models. Although 6-hydroxydopamine has been and continues to be the most commonly used technique for causing degeneration of nigrostriatal dopamine neurons, there is no question that 1-methyl-4-phenyl-1,2,5,6-tetrahydropyridine (MPTP) has led to an improved understanding of basal ganglia circuitry and has generated valuable treatment strategies through primate studies. We can all agree that breakthroughs derived from one species should not automatically preclude the usefulness of other species for further discoveries. However, rather than focusing on whether one species might be more relevant than another as a model for investigating neurological disorders and developing treatments, the aim of our paper was to suggest that the way in which a species is used cannot be ignored. Inadequate behavioural measures in rats, or in primates, can undermine any advantages a research project might offer.