3Rs implementation

In 1959, two British academics, William Russell and Rex Burch, members of the University Federation for Animal Welfare—an association that remains highly active in the field of animal welfare—published a book destined to become a cornerstone in the evolution of animal experimentation: The Principles of Humane Experimental Technique. In this volume, the two scholars presented an approach to experimental research involving animals that took into account both data quality and the welfare of the animals used. This formula is known as the “3Rs” principle or model: Replacement, Reduction, and Refinement.

Anyone who needs animals for research must first ask whether it is possible to replace the chosen animal model with methods that achieve the same result without using animals (Replacement). If this is not possible, every means must be employed to minimize the number of individuals used, without compromising the reliability of the results (Reduction).

Finally, the animals used—regardless of species—must be treated with all measures most suitable to make procedures as minimally impactful on their welfare as possible, reducing as far as possible any kind of suffering that may result from experimental procedures (Refinement). The “3Rs” have evolved over time, although their underlying principles have remained the same, and they are now formally integrated into European and national legislation protecting the welfare of animals used for scientific purposes.

Replacement
Declaration that the animal species that will be used are those with the lowest neurological development, as well as the lack of alternative methods, compatible with the objective of the research project, Legislative Decree 26/2014, art. 13 paragraph 2.

Ongoing basic research projects aim to understand the neuronal mechanisms underlying high-level motor and socio-cognitive functions, expressed in the behavioral repertoire of only a few primate species beyond humans.
The first and most obvious option to replace the animal model would therefore be to study the same functions directly in human subjects. This is indeed possible and widely pursued using indirect, non-invasive techniques to investigate brain activity, such as functional magnetic resonance imaging (fMRI), high-density electroencephalography (EEG), transcranial magnetic stimulation (TMS), and magnetoencephalography (MEG). Moreover, over the past 20 years, the use of prolonged intracranial EEG monitoring for pre-surgical localization of epileptogenic foci in patients with drug-resistant epilepsy and negative MRI findings has enabled more detailed investigations of human brain function with invasive techniques employed primarily for clinical purposes (see, for example, 1). Computational simulation models offers a further means to integrate acquired data and to formulate empirically grounded hypotheses and predictions, but it requires direct experimental data. These approaches have contributed to a marked reduction in the use of animals – particularly non-human primates – in neuroscience research (from ~8,000 to ~6,000 in Europe between 2008 and 2011, −25%; source: SCHEER report 2017). However, none of these techniques currently allows replacement of the animal model for studies at the single-neuron and neural-network levels, which require direct electrophysiological recordings.

Studies involving simultaneous recordings from multiple regions are essential to bridge the unit level (single cells) to the macro-system (networks) and to draw causal inferences about the mechanisms that generate behavior. This paradigm, already widely developed in rodent models, is not transferable to our research questions when fine motor repertoire, communication, and social interaction demand anatomical–functional and behavioral homologies that only non-human primates provide. In this context, macaque monkeys represent the least neurologically complex species suitable for achieving the projects’ objectives.

Macaque monkeys (notably M. fascicularis and M. mulatta) are reference models in biomedical and neuroscience research⁴, with the anatomical–functional, cognitive, and behavioral homologies with humans necessary and sufficient to yield results of comparative and evolutionary relevance⁵. For example, forelimb movement abilities and the fine use of the hands for reaching and manipulating objects rely on evolutionarily conserved mechanisms and substrates⁶. Macaques also display a broad social and affiliative repertoire – expressed through postures, facial and gestural signals, vocalizations, and their multisensory combinations – that is not comparable to species with less developed nervous systems⁷.

Macaca mulatta

Macaca fascicularis

More specifically, the species Macaca mulatta is widely used in behavioral neurophysiology studies due to its excellent adaptability and tolerance to housing and laboratory conditions, if appropriate environmental enrichment and social housing measures are in place⁸. It is also the species with which our group has the greatest direct experience: its biology and ethology have been investigated in numerous neurophysiological research projects at the University of Parma, authorized by the Italian Ministry of Health (Authorizations 128/2005-B, 129/2005-B of 20/09/2005; 60/2006-C, 61/2006-B of 20/04/2006; 207/2008-C of 24/11/2008; 54/2010-B, 55/2010-C of 18/03/2010; 294/2012-C of 11/12/2012), and further consolidated through specific training for personnel performing the tasks and functions referred to in Article 23(2) of Legislative Decree 26/2014 (Ministerial Decree of 5 August 2021, module specific to non-human primates) and ongoing collaborations with the German Primate Center in Göttingen, where Macaca mulatta is the most widely used species in neuroscience.

Reduction
Declaration that the animal species that will be used are those with the lowest neurological development, as well as the lack of alternative methods, compatible with the objective of the research project, Legislative Decree 26/2014, art. 13 paragraph 2.

Non-human primate neuroscience studies predominantly adopt within-subject designs: sample size is defined not by the number of animals, but by the number of independent measurements reliably and reproducibly obtained from each individual. To address inter-individual variability and ensure robustness and reproducibility, neurophysiology literature sets as a standard the replication of the effect in a second animal. Consequently, the minimum number of animals necessary and sufficient for each neurophysiological experiment is typically two.

The use of new technologies that allow the same animals to participate in different experiments (e.g., injections of pharmacologically active substances and neuroanatomical studies in the same subject) increases the quality and completeness of results while reducing the overall number of animals needed for an equivalent information yield, without increasing cumulative severity.

Refinement
Optimization of the method to reduce the suffering imposed on the animal during the execution of the procedures, Legislative Decree 26/2014, art. 13 paragraph 2.

Refinement in research on non-human primates is both an ethical and legal obligation and an essential means of optimizing animal compliance and maximizing data quality (see Refinement techniques in non-human primate neuroscientific research). All procedures—from housing to surgical interventions and experimental sessions—are conducted according to the highest international standards. Our effort to definitively move beyond head-fixation systems in favor of wireless technologies with freely moving animals represents a step change that goes beyond current standards and implements the most recent European recommendations (cf. SCHEER report 2017).

Animals are sourced from authorized European suppliers who guarantee the availability of captive-born animals from captive-born parents (F2—neither the animals nor their parents are ever taken directly from the wild). Transport is carried out by trained and authorized personnel using appropriate means. From acquisition onward, with a view to improving cumulative lifetime experience, we apply refinement measures that cover all management aspects—from housing to experimental procedures—as detailed in the following sections.

▪ Training of the experimenters

The assessment, monitoring, and optimization of the environment and procedures throughout the animals’ entire life cycle can yield substantial benefits only insofar as they are carried out by personnel who are experienced and appropriately trained for this purpose. For this reason, all staff with direct responsibility for procedures or for the animals’ daily care receive specific theoretical–practical training, demonstrated by completion of the training required for personnel performing the tasks and functions referred to in Article 23 comma 2 of Legislative Decree No. 26/2014 (Ministerial Decree of 5 August 2021) specific to non-human primates before initiating any interaction with the animals. Ongoing exchange with colleagues and veterinarians working in other European primate centers is part of continuous education and training and complements the staff’s periodic updates through participation in seminars and events specifically focused on the 3Rs and animal welfare.

▪ Housing and enrichments

Our facility adopts an interconnected cage system compliant with Directive 2010/63/EU and Legislative Decree 26/2014, designed and manufactured by a leading company in collaboration with major European primate centers. Animals are housed in pairs or small groups in spaces larger than the legal minimum (≥ 2.5 m³ vs. 1.8 m³ required for each animal).

Environmental enrichment follows a daily rotation schedule that includes physical/structural items (lianas, suspended toys, climbing elements), sensory stimuli (mirrors, visual and/or auditory stimuli), feeding-related enrichment (sawdust or natural bark substrates for dispersing seeds and small food items; perforated objects for foraging), cognitive-occupational activities (puzzles and problem-solving tasks), and social enrichment. Two recreation cages equipped with wooden structures and swings are also available and accessed by animal pairs on a daily rotation, allowing them to maintain and exercise their full species-typical behavioral repertoire (e.g., climbing, jumping, exploration). The facility also includes a group-housing room (L × W × H: 3.43 × 2.30 × 2.72 m), equipped with a swing, a large wooden structure for climbing and jumping, natural bark flooring, and a direct connection to other cages via a tunnel.

A sound system and video monitors provide additional sensory and cognitive enrichment.

The environment benefits from large windows for optimal natural light and regular day/night cycles; artificial lighting is controlled by timers. The microclimate is managed by a temperature-control system with 24/7 active alarms.
Sanitary management of the facility and daily nutrition are handled by trained technical staff, who operate according to the indications and under the supervision of the research staff and the designated local animal-welfare officer.

▪ Positive reinforcement learning

All animals are trained exclusively with methods based on positive reinforcement and voluntary cooperation. The process begins with a phase of gradual familiarization with the experimenters and equipment, during which simple commands are delivered through the cage and reinforced with small food rewards (e.g., fruit), distinct from the animals’ daily dry ration.

Habituation procedures and task-specific training are carried out by competent personnel using operant-conditioning techniques based on positive reinforcement¹¹. Clicker training is systematically employed as an immediate auditory marker of the correct behavior that precedes the reward, thereby facilitating and accelerating learning^12,13.

For experimental sessions, animals are progressively trained to approach, enter, and remain in the primate chair for initially brief and then progressively longer periods, without the use of collars or rigid spacers and restraints^14,15. Sessions are scheduled with gradual objectives, kept within a limited duration, and are interrupted or adjusted in the presence of signs of stress, in order to safeguard animal welfare and data quality.

In our laboratory, we are systematically adopting procedures that eliminate head-fixation devices and implement the use of transparent plexiglass transfer carriers to move animals into the environment equipped for wireless recordings. This approach enables shorter habituation timelines (typically a few weeks) with minimal signs of stress or refusal, compared with traditional protocols that may require up to a year – particularly for the delicate phase of habituating animals to repeatedly accept limited mobility in the primate chair. Gradual progression remains a guiding principle: training proceeds stepwise, thereby reducing the risk of regression and fostering voluntary cooperation.

Task-specific training for experimental paradigms is conducted using the same evidence-based behavioral methodologies: task analysis, decomposition of behavior into elementary components, differential reinforcement of spontaneous behaviors that approximate the target, response shaping, and cue fade-out used as intermediate steps. The goal is to select and reinforce behaviors closest to the desired behavior, promoting the extinction of alternatives and leading to the progressive acquisition of the required skills even in cognitively complex tasks. This approach – fully aligned with best practices used by leading professional dog trainers – constitutes a Refinement within the 3Rs, combining animal welfare with data quality.

▪ Miniaturization and refinement of invasive devices

All implanted devices (such as the interconnection system for the wireless transmitter) are made of biocompatible materials and minimized in size. The multielectrode probes used typically have diameters between 50–100 µm, thereby limiting tissue damage to negligible levels, as confirmed by histological analyses in previous studies¹⁵.

Image courtesy of Microprobes for Life Science and Plexon.

Head-fixation devices still in use, employed only where necessary, have proven to be particularly well tolerated, showing excellent osseointegration, as also evidenced by post-mortem examinations¹⁶. In line with the Refinement principle of the 3Rs, we are steadily transitioning to a non-invasive, fully external, and temporary head-immobilization system that does not require any fixed cranial implant. This support is applied only for a few moments, solely to stabilize the head while connecting the wireless recording system, and is then removed. The approach reduces animal handling, shortens pre-session preparation, and mitigates the risk of infectious or inflammatory complications associated with the presence of a chronic implant.

In parallel, we adopt, whenever feasible, chronic or semi-chronic recordings with the cranial bone left intact over the region of interest, along with protocols that leave no residual discomfort after probe implantation (performed under general anesthesia). Overall, these choices minimize handling before each session, drastically reduce the risks of infection or inflammation, and further enhance animal welfare.

▪  Anesthetics, surgical techniques and post-operative pain medication.

All major surgical procedures are performed in the presence of the Designated Veterinarian, under inhalational general anesthesia with halogenated agents, administered by a veterinarian with specific certification required for personnel performing the tasks and functions referred to in Article 23 comma 2 of Legislative Decree 26/2014 (Ministerial Decree of 5 August 2021, module specific to non-human primates).
Intraoperative monitoring of temperature, respiratory rate, spirometry, oxygen saturation, heart rate, ECG tracing, and non-invasive arterial blood pressure is ensured by a multiparameter monitor operated by the anesthetist. A neurosurgeon is available for oversight and updates and takes part in all procedures where needed.

For preoperative planning and intraoperative targeting, the laboratory employs a neuronavigation system that integrates MRI and CT to create subject-specific models and support the definition of optimal trajectories and guided execution of procedures. This enables accurate, targeted implantations, helping to reduce time and handling and to improve safety and reproducibility.

For preoperative planning and intraoperative targeting, the laboratory employs a neuronavigation system that integrates MRI and CT to create subject-specific models and support the definition of optimal trajectories and guided execution of procedures. This enables accurate, targeted implantations, helping to reduce time and handling and to improve safety and reproducibility.

The postoperative course is overseen by the animal-care staff under the supervision of the Designated Veterinarian and the anesthetist, who prescribe and adjust antibiotic, anti-inflammatory, and analgesic therapy until full recovery.

▪  Recording techniques that maximize the quantity and quality of collected data, reducing the number and duration of recording sessions

A further refinement involves optimizing the use of MRI under anesthesia to obtain high-resolution images of skull and brain morphology. The 3D reconstruction of a plastic model of each animal’s skull makes it possible to anatomically conform the implanted devices to the specific bone structure: results obtained to date show lighter, less invasive implants that require no maintenance, disinfection, or cleaning.

The study of brain morphology allows precise identification of the region of interest and underpins the use of chronic or semi-chronic probes implanted exactly in that region.

Simultaneous multichannel recording enables the acquisition of activity from hundreds of sites in parallel using only a few high-density probes. Behavioral parameters—such as movement kinematics or eye position—are monitored with fully non-invasive methods based on IR cameras and dedicated infrared light sources, ensuring signal stability under any visible-light conditions.

▪  Use of the same animals for neurophysiological and neuroanatomical studies, without impact in terms of “cumulative-severity”.

In our laboratory, the same animals can be used for both neurophysiological and neuroanatomical studies without an increase in cumulative severity. The implantation of intracortical probes is performed with miniaturized devices and, in itself, does not entail significant damage or suffering for the animal. When the goal is to define the connectivity of regions investigated electrophysiologically, we use injectrodes – silicon probes equipped with one or more microchannels for the infusion of microliter volumes of pharmacologically active substances or, at the end of the experiments, neuronal tracers. These tracers, with retrograde or retro-anterograde transport, allow – after an interval appropriate to the tracer—the post-mortem verification of the anatomical connectivity of the same region studied functionally.

Neuroanatomical studies do not require procedures beyond those already planned for neurophysiological studies and do not require dedicated animals; consequently, they do not produce additive effects in terms of cumulative severity. When it is scientifically necessary to reconstruct with precision the anatomical network of the studied brain, painless euthanasia is carried out using species-appropriate methods. When such verification is not required or is not part of the experimental plan, animals are enrolled in a rehoming program at accredited rescue centers, in accordance with current regulations and following veterinary assessment (health suitability, observation period, behavioral requirements).

Current MR tractography techniques remain insufficient in terms of resolution, reliability, and reproducibility, and do not represent a comparable alternative to tracer studies. As soon as advances in non-invasive tractography provide data of adequate quality – or in cases where reconstructing the anatomical network in the same subject is not necessary – we will consider rehoming the animals to rescue centers at the end of the experiments.

▪  Continuous monitoring of animal welfare

Despite the measures described above, unpredictable adverse events may occur that affect an animal’s welfare (e.g., accidental trauma, injuries from conflict with a partner, transient infections, or organic dysfunctions unrelated to experimental procedures). Such events can arise independently of research activities, as can happen with any domestic animal.

To prevent and effectively manage potential negative effects, we implement continuous welfare monitoring carried out by trained and qualified personnel. For each subject, an individual clinical assessment sheet is completed upon entry into the facility and updated daily throughout the animal’s life cycle. Daily monitoring records information that is broader than what is required by the legally mandated personal history file: clinical and behavioral observations; quantity and type of food and fluids consumed; body weight; noteworthy events or behaviors; type of enrichment provided and degree of interaction; performance during training and any experimental tasks.

All animals involved in a procedure – even those only in the training phase – are observed or interact each day with at least one designated experimenter or technician, who can promptly detect any deviation from optimal status and initiate appropriate corrective actions, such as veterinary evaluation, modification or pause of training, or adaptation of environmental conditions or enrichment. The information collected enables protocols to be adjusted and optimized according to each animal’s individual characteristics.