Banca de QUALIFICAÇÃO: RAFAEL DOS SANTOS DE BESSA

Uma banca de QUALIFICAÇÃO de DOUTORADO foi cadastrada pelo programa.
STUDENT : RAFAEL DOS SANTOS DE BESSA
DATE: 31/08/2022
TIME: 14:00
LOCAL: https://meet.google.com/gaw-tshj-qzj
TITLE:

THE ONTOGENY OF ULTRASONIC VOCALIZATIONS IN AN ANIMAL MODEL OF AUTISM

KEY WORDS:

Autism; USV; Postnatal development; GMM; Unsupervised learning;

PAGES: 140
BIG AREA: Ciências Biológicas
AREA: Biologia Geral
SUMMARY:

Communication during social interaction is one of the main factors for the best behavioral response to each situation. Rats use ultrasonic vocalizations (USVs) to communicate with other conspecifics. The vocalizations carry information about the rat's emotional state and contribute to the start and maintenance of social behaviors. Neonate rats emit USVs when they are far away from their mother and in stressful situations. The so-called 40-kHz USVs evoke an approximation response of the mother to the pup and maternal care. As the animal age, the USVs change their acoustic profile and become closer to that observed in adults. Understanding how the acoustic parameters change can help to identify biomarkers of development and disorders in which communication and social interaction are affected as in the autism spectrum disorder (ASD). ASD patients show communication and social interaction impairments early in life. Animals exposed to valproic acid (VPA) in the uterus have anatomical and behavioral alterations similar to those observed in human ASD patients. These animals show a reduction in the number of USVs emitted during childhood maternal isolation. However, little is known about how the acoustic parameters are affected in this model. In this study, we analyzed the acoustic parameters of USVs emitted by neonate rats during maternal isolation on postnatal days 07, 14, and 21. We also analyzed the effect of VPA exposition in the uterus on vocal production. To qualitatively evaluate the USVs we developed an unsupervised clustering algorithm that applies a gaussian mixture model (GMM) and uses only two USVs parameters: the principal frequency and the frequency modulation. We analyzed the transition dynamics between the different identified clusters and tested for existing correlations between the proportion of each USV class and exploratory behaviors observed during an open field test on postnatal day 21 (P21). Our results showed that VPA-exposed animals vocalize less than control animals at P14 and their USVs show altered acoustic parameters at all recorded ages. Our algorithm identified four USV clusters/classes. We found differences in the proportion of USV classes emitted by VPA animals and controls on P21 and showed that VPA animals emit fewer USV classes at this age, which means a less diverse vocal repertoire. The analyses of exploratory behaviors in the open field showed that VPA animals explore less than control animals. We also showed that the animals which emitted more USV classes with higher frequency modulation (loFM and hiFM) had more exploratory behaviors. Animals that emitted more USV classes with lower frequency modulation (loNM and hiNM) had fewer exploratory behavior in the open field. These results show that our algorithm is capable of separating neonate USVs into four classes using only two acoustic parameters. The correlations found between the USV classes and the exploratory behaviors, and the differences in the proportion of these classes between VPA animals and controls indicate a biological correspondence of these classes and open the possibility of using the 40-kHz USvs as biomarkers.

COMMITTEE MEMBERS:
Interno - 1996111 - MARTIN PABLO CAMMAROTA
Interno - 1824636 - RICHARDSON NAVES LEAO
Presidente - 1698305 - RODRIGO NEVES ROMCY PEREIRA