UNIVERSIDAD DE COSTA RICA 
SISTEMA DE ESTUDIOS DE POSGRADO 

 
 
 
 
 
 
 
 

CARACTERIZACIÓN MOLECULAR Y FENOTÍPICA DE LA ATAXIA CON 
APRAXIA OCULOMOTORA EN COSTA RICA MEDIANTE ENFOQUES         

MOLECULARES Y BIOINFORMÁTICOS 
 
 
 
 
 
 
 
 
 
 

Tesis sometida a la consideración de la Comisión del Programa de Estudios de         
Posgrado en Biología para optar al grado y título de Maestría Académica en Biología 

con énfasis en Genética y Biología Molecular 
 

 
 
 
 

HILDA GUADALUPE TORRES ULATE 
 
 
 
 
 
 
 
 
 

Ciudad Universitaria Rodrigo Facio, Costa Rica 
 
   

2022 



 
 

 

ii 
 

 
 
 

Agradecimientos 

 
 
 

 

Para mis padres por su apoyo, consejos, ayuda en los momentos difíciles, y por brindarme 

los recursos necesarios para estudiar. Me han dado todo lo que soy como persona, mis 

valores, mis principios, mi carácter, mi empeño, mi coraje para conseguir mis objetivos 

y no desmayar en los problemas que se presentaban, enseñándome a encarar las 

adversidades sin perder nunca la dignidad ni desfallecer en el intento. A mis hermanos 

por su compañía y sabias palabras en los momentos necesarios. 

 

Agradezco también a mi comité asesor. A Ph.D. Alejandro Leal Esquivel, Ph.D. Gabriela 

Chavarría Soley y a Ph.D. José Guevara Coto por toda la colaboración y el apoyo. Gracias 

por su enseñanza durante este proceso.  

 

A, Martha Gómez, Saby Cruz, Lidia Benítez, Ricardo Villalobos, Marcelo Castro, María 

de los Ángeles Ulate, Monserrat Lobo y Carolina Céspedes por toda la ayuda, compañía, 

el aliento brindado y por sus valiosos aportes para el desarrollo de esta investigación. 

 
 
 
 
 
 
 
 
 
 
 



 
 

iii 
 

Hoja de aprobación 

“Esta Tesis fue aceptada por la Comisión del Programa de Estudios de Posgrado en 
Biología de la Universidad de Costa Rica, como requisito parcial para optar al grado y 

título de Maestría Académica en Biología con énfasis en Genética y Biología Molecular 
 
 
 
 
 

Dr. Andrey Sequeira Cordero 
Representante de la Decana 

Sistema de Estudios de Posgrado 
 
 
 
 
 

Dr. Alejandro Leal Esquivel 
Profesor Guía 

 
 
 
 
 

Dra. Gabriela Chavarría Soley 
Lectora 

 
 
 
 
 

Dr. José Andrés Guevara Coto 
Lector 

 
 
 
 
 

Dr. Eric Fuchs Castillo 
Director del Programa de Posgrado en Biología 

 
 
 
 
 

Hilda Guadalupe Torres Ulate 
Sustentante 

 



 
 

iv 
 

Contenido 

 

Contenido 
Agradecimientos ....................................................................................................... ii 

Hoja de aprobación ................................................................................................. iii 

Contenido ................................................................................................................. iv 

Resumen .................................................................................................................... v 

Lista de Cuadros ..................................................................................................... vii 

Introducción .............................................................................................................. 1 

Metodología .............................................................................................................. 8 

Objetivo General ..................................................................................................... 15 

Objetivos Específicos ............................................................................................. 15 

Artículo: Analysis of ataxia in Costa Rica through an integrative approach .......... 16 

Conclusiones ........................................................................................................... 38 

Referencias Bibliográficas ...................................................................................... 40 

 

 

 

 

 



 
 

v 
 

Resumen 

 
Las ataxias son un grupo de trastornos clínicamente heterogéneos, que pueden ser here-

dables o no dependiendo de la etiología. Las manifestaciones clínicas de las ataxias he-

reditarias son una incoordinación progresiva del movimiento y del habla (disartria), y una 

marcha inestable, descoordinada y de base amplia. Además, los pacientes pueden desa-

rrollar oftalmoplejía (limitaciones del movimiento ocular), espasticidad, neuropatía y di-

ficultades cognitivas. Este trabajo se enfoca en la ataxia con apraxia oculomotora (AOA), 

que es una enfermedad de herencia autosómica recesiva. Hasta la fecha se han descrito 

cuatro tipos de AOA (AOA1-AOA4). Los diferentes tipos de ataxia comparten varias de 

sus características clínicas, los cuales se diagnostican mediante la historia familiar, el 

examen físico, la neuroimagen y las pruebas moleculares. Debido a que la información 

sobre las mutaciones causantes de AOA en Costa Rica es escasa, se planea identificar 

molecularmente las mutaciones causantes de AOA en pacientes costarricenses mediante 

los métodos de secuenciación de Sanger y secuenciación de nueva generación, por lo que 

este trabajo contribuiría a caracterizar las mutaciones prevalentes en los pacientes que 

padecen ataxia con apraxia oculomotora en el país, así como el fenotipo asociado. Ade-

más, sentaría las bases para desarrollar protocolos de trabajo para el diagnóstico molecu-

lar de la enfermedad.  

El desafío de determinar si las manifestaciones fenotípicas están asociadas con grupos 

específicos de proteínas que comparten una relación subyacente, representa una oportu-

nidad para estudiar cómo surgen las diferentes manifestaciones fenotípicas. Pero además, 

los enfoques bioinformáticos recientes en el estudio de las proteínas y las relaciones entre 

aminoácidos podrían permitir el análisis de agrupamiento e identificación de grupos de 

residuos involucrados en los diferentes tipos de AOA. Esto con el objetivo de asociar los 



 
 

vi 
 

fenotipos clínicos reportados a grupos de residuos proteicos que se encuentran conserva-

dos o relacionados con otros residuos dentro de la misma estructura de la proteína. 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



 
 

vii 
 

 

 

 

 

 

Lista de Cuadros 

Cuadro 1. Resultados de pruebas serológicas y sus variaciones según el tipo de AOA …3 

Cuadro 2. Mutaciones descritas e incluidas en OMIM hasta la fecha, para cada uno de los 

tipos de AOA …………………………………………………………………………...6 

  



 
 

 

 

 

 



 
1 

 

 

 

Introducción 

 
Las ataxias son un grupo de trastornos dentro de las enfermedades cerebelares, de las 

cuales es posible encontrar registros que datan de mediados desde el siglo XIX y 

principios del siglo XX cuando se hicieron los primeros esfuerzos por tratar de 

clasificarlas (Holmes, 1908). Estos trastornos clínica y genéticamente heterogéneos, se 

caracterizan por una descoordinación lentamente progresiva de la marcha. A menudo son 

asociadas con una mala coordinación en el movimiento de las manos, movimientos 

oculares, habla, y con frecuencia se produce la atrofia del cerebelo. Las ataxias pueden 

ser:  

a. No hereditarias: cuando aparecen en la adultez (Klockgether, 2010) son causadas 

por factores tales como alcoholismo crónico, deficiencias vitamínicas, enferme-

dad vascular, atrofia muscular espinal, etc. (Jayadev & Bird, 2013).  

b. Hereditarias: cuando son causadas por una mutación causal en un gen. Hasta la 

fecha, se sabe que hay más de 30 formas autosómicas dominantes y más de 60 

formas que son autosómicas recesivas (Ruano, Melo, Silva, & Coutinho, 2014). 

Muchas de las ataxias hereditarias tienen presentaciones superpuestas y existe un 

alto grado de heterogeneidad genética (Inlora et al., 2017).  

Las manifestaciones clínicas de las ataxias hereditarias son una incoordinación 

progresiva del movimiento y del habla (disartria), y una marcha inestable, descoordinada 

y de base amplia. Además, los pacientes pueden desarrollar limitaciones del movimiento 

ocular (oftalmoplejía), espasticidad, neuropatía y dificultades cognitivas (Jayadev & 

Bird, 2013). Para establecer el diagnóstico de ataxia hereditaria se requiere:  

• Detectar signos clínicos típicos en el examen neurológico (Klockgether, 2010).  



 
2 
 

 

• Documentación de la posible naturaleza hereditaria de la enfermedad mediante el 

desarrollo de un historial familiar (Jayadev & Bird, 2013). En caso de no poder 

identificar una causa genética es necesario proceder con la exclusión de causas no 

genéticas (Klockgether, 2010).  

 

Dentro de las ataxias de herencia autosómica dominante, Schöls et al; 2004 definen a las 

ataxias espinocerebelosas (SCA) como un grupo clínica y genéticamente heterogéneo que 

causa la degeneración progresiva del cerebelo y sus conexiones aferentes y eferentes. 

Estos trastornos están causados por una mutación de expansión de repetición del triplete 

CAG en regiones codificantes de los genes. (Pulst et al., 1996). Las SCA más comunes 

son SCA1, SCA2, SCA3 y SCA6 (Schmitz-Hübsch et al., 2008). Esto puede variar en 

ciertas regiones debido a un efecto fundador (Manto, 2005).  

 

Los pacientes suelen presentar un síndrome cerebeloso lentamente progresivo con 

diversas combinaciones de trastornos oculomotores, disartria, dismetría/temblor cinético 

y/o marcha atáxica. Al ser  genéticamente heterogéneas puede presentarse un traslape de 

los fenotipos entre los diferentes subtipos (Manto, 2005). 

 

Durante los últimos quince años se han descrito el segundo grupo de ataxias cerebelosas 

con la mayor prevalencia a nivel mundial dentro de las ataxias con herencia autosómica 

recesiva (Ruano et al., 2014).  

Estas ataxias son causadas por mutaciones en genes responsables de reparación del ADN, 

terminación prematura de la proteína, deficiencias de maduración de la proteína, o ambos. 

Dentro del mismo se incluyen la ataxia con apraxia oculomotora (AOA), ataxia 



 
3 
 

 

telangiectasia (AT) y el más inusual, el trastorno similar a ataxia telangiectasia (ATLD) 

(Mariani et al., 2017). 

Con respecto a la AOA, hasta la fecha se han descrito cuatro tipos: AOA1, AOA2, AOA3 

(Tassan et al., 2012) y AOA4/CMT2B (Leal et al., 2018). Los diferentes tipos de AOA 

comparten síntomas entre ellos (Inlora et al., 2017). 

La AOA1 tiene una edad de inicio que va desde la infancia hasta la preadolescencia (2-

12 años) y es causada por mutaciones en el gen APTX (Coutinho & Barbot, 2002) (Cuadro 

1), que se localiza en 9p21.1; tiene 7 exones y un tamaño de 28.217 pb. (Kent et al., 

2002).  APTX codifica para la proteína aprataxina, que se encarga de catalizar la 

liberación nucleofílica de los grupos adenilato unidos covalentemente a los extremos 5’ 

fosfato en rupturas de una sola hebra, dando como resultado la producción de los 

extremos 5’ fosfato que se pueden unir de manera eficiente durante la reparación del 

ADN (MIM 2089200; Prasad et al., 2009). 

Cuadro 1. Mutaciones descritas e incluidas en OMIM hasta la fecha, para cada uno de 
los tipos de AOA 

AOA1 
APTX 

AOA2 
SETX 

AOA3 
PIK3R5 

  AOA4 
PNKP 

● p.Lys247 
● p.Pro206Leu 
● 689insT 
● 318delT 
● p.Val89Gly 
● p.His27Arg 
● Trp279Ter 
● p.Trp279Arg 
● Del.  7 exones 
● Leu223Pro 
 c.689dupT 
 p.Glu232Gly-

fsTer38 

● c.839-2A>G 
● p.Thr2154Met 
● p.Arg168Trp 
● p.Leu1976Arg 
● p.Glu65Lys 
 

●  p.Pro629Ser 
 

● c0.1189–15_1191del18 
●  p.Gly375Trp 
● Gly442AlafsTer27 
● Thr408del 
● Gln517LeufsTer24 

 

Fuente: Gatti et al., 2019; Laurencin et al., 2015; MIM. 2089200; Duquette et al., 2005; Mignarri, Tessa, 
Federico, Santorelli, & Dotti, 2015; Szpisjak, Obal, Engelhardt, Vecsei,  Klivenyi, 2016; Tassan et al., 2012; 
Bras, et al., 2015 and Schiess et al., 2017. 
 



 
4 
 

 

La aprataxina es un miembro de la superfamilia de la triada de histidinas (HIT) (van 

Minkelen et al., 2015), que son una antigua superfamilia de nucleótido hidrolasas y 

transferasas que actúan sobre el α-fosfato de ribonucleótidos o en sustratos que contienen 

nucleótidos en vías de señalización importantes para el crecimiento celular, apoptosis y 

metabolismo de ADN, ARN y carbohidratos (Brenner, 2002). La aprataxina se expresa 

en los siguientes tejidos: cerebro, tiroides, paratiroides, glándulas adrenales, pulmones, 

páncreas, hígado, vesícula biliar, vejiga, corazón, riñón, hígado, pulmón, linfa, 

mioblastos, miotubos, páncreas, placenta, músculo esquelético, cordón espinal y tálamo 

(Uhlen et al., 2015).  

La AOA2 se describió por primera vez en el año 2000 (M. C. Moreira et al., 2004). Es 

ocasionada por mutaciones en el gen SETX (Cuadro 3) y tiene una edad media de inicio 

en la pre adolescencia (12.7 años) (Schiess, et al., 2017). Codifica para una proteína que 

por su homología con la proteína Sen1p de los hongos, se denominó Senataxina. La 

senataxina tiene actividad de ARN helicasa codificada por un dominio helicasa de 

ADN/ARN en el extremo C-terminal, lo que sugiere que puede estar involucrada en el 

procesamiento de ADN y ARN (Prasad et al., 2009).  Se expresa en: músculo esquelético, 

corteza cerebral, apéndice, testículos, músculo cardíaco, cerebelo, músculo liso, pulmón, 

nasofaringe, intestino delgado, hipocampo, piel, tiroides, endometrio, médula ósea, 

bronquios, riñón, recto, trompas de Falopio, estómago, epidídimo y colon. (Uhlen et al., 

2015) 

Tassan y colaboradores (2012) identificaron la AOA3 en una familia consanguínea de 

Arabia Saudita, cuyos miembros afectados presentaban características clínicas similares 

a las de los individuos con AOA2, pero con una edad media de inicio en la adolescencia 

(15.6 años). Este tipo de ataxia es causada por mutaciones en el gen PIK3R5 (Tassan et 

al., 2012) (Cuadro 3), que se localiza en 17p13.1, tiene 18 exones y un tamaño de 30.869 



 
5 
 

 

pb (Kent et al., 2002). Este gen codifica por la subunidad reguladora 5 del complejo de 

clase fosfatidilinositol 3-quinasas gamma (PIK3γ), que es una subunidad reguladora de 

101 kD del complejo de clase I PIK3γ. PIK3R5 es un enzima dimérica (Uhlen et al., 

2015). Las fosfatidilinositol 3-quinasas (PI3K) son miembros de una familia única y muy 

conservada de quinasas intracelulares que fosforilan el grupo 3'-hidroxilo del 

fosfatidilinositol y fosfoinosítidos. Esta reacción tiene como resultado, la activación de 

múltiples vías de señalización intracelular que regulan funciones diversas e importantes, 

como el metabolismo celular, la supervivencia y la polaridad, y el tráfico de vesículas 

(Engelman, Luo, & Cantley, 2006). Esta proteína se expresa en duodeno, estómago, 

apéndice, pulmón, nasofaringe, vesícula biliar, tiroides, glándula adrenal, intestino 

delgado, colon, recto, riñón, epidídimo, vesícula seminal, cérvix, bronquios, placenta, 

glándulas salivales, trompas de Falopio y médula ósea (Prasad et al., 2009). 

El último tipo descrito de AOA presenta una edad de inicio promedio de la enfermedad 

en la infancia (4.3 años) (Schiess et al., 2017) y en algunos casos se caracteriza por ataxia 

cerebelosa, apraxia oculomotora, polineuropatía y atrofia cerebelosa en la resonancia 

magnética, dicho fenotipo se clasificó como ataxia con apraxia oculomotora tipo 4 

(AOA4) (Gatti et al., 2019). En otros casos, el fenotipo clínico presentado se caracterizó 

por la presencia de una polineuropatía motora y sensorial parecido a una forma de la 

enfermedad de Charcot-Marie-Tooth e identificado como CMT2B2 (Leal et al., 2018; 

Pedroso et al., 2015). La causa se debe a mutaciones en el gen PNKP (Cuadro 3), el cual 

se ubica en 19q13.33, cuenta con 16 exones y tiene un tamaño de 6.337 pb. PNKP 

codifica por una fosfatasa polinucleótido quinasa 3’, que en respuesta al daño por 

radiación ionizante o al daño oxidativo del ADN, cataliza la fosforilación 5’ de ácidos 

nucleicos y también tiene una actividad fosfatasa 3’ asociada, que predice una función 

importante en la reparación del ADN (Uhlen et al., 2015). Se ha reportado la expresión 



 
6 
 

 

de esta proteína en cerebro, colon, corazón, riñón, hígado, pulmón, ovarios, páncreas, 

placenta, próstata, músculo esquelético, intestino delgado, bazo y testículos (Prasad et 

al., 2009). 

Debido a que los diferentes tipos de ataxia comparten varias de sus características clínicas 

(Mariani et al., 2017), estos se diagnostican mediante la historia familiar, el examen 

físico, marcadores serológicos (Cuadro 2), la neuroimagen y las pruebas moleculares, las 

cuales permiten detectar mutaciones nuevas o previamente descritas, como las que se 

muestran en el Cuadro 3.  

Cuadro 2. Resultados de pruebas serológicas y sus variaciones según el tipo de AOA 

AOA 1 AOA 2 AOA 3 AOA 4 
● Albúmina disminuida  
● Colesterol total aumen-

tado 
● Alfa fetoproteína normal 
● Deficiencia de coenzima 

Q10 a nivel muscular 
puede ayudar, pero no es 
definitiva.  

● Conteo sanguíneo, 
creatinquinasa y coles-
terol normales 

● Alfa fetoproteína ele-
vada  

 

● Alfa fetoproteína 
elevada  

 

● Albúmina elevada 
● Colesterol elevado 
● Alfafetoproteína normal. 
 

 

Fuente: van Minkelen et al., 2015; Becherel, et al., 2015; Tassan et al., 2012 and Schiess, 
Zee, Siddiqui, Szolics, & El-Hattab, 2017 
 

En cuanto a los resultados de pruebas serológicas, se encuentran diferencias en los 

biomarcadores listados en el Cuadro 3, sin embargo, estos no pueden utilizarse como una 

factor discriminativo ya que como ha sido reportado por Coutinho y Barbot (2015) así 

como por Mignarri, et al. (2015), los resultados de los marcadores serológicos pueden 

traslaparse en los diferentes tipos de AOA, e incluso se han reportado casos de individuos 

con un mismo tipo de AOA y resultados variables en los biomarcadores serológicos.  

Hasta ahora no se han estudiado las AOAs y su causa genética en Costa Rica.  No 

obstante, Leal et al (2018) reportaron las mutaciones p.Gln517ter, en condición 

homocigota y junto con Thr408del en el gen PNKP en pacientes heterocigotos 

compuestos, como responsable de Charcot Marie Tooth (CMT) con afectación cerebelar.  



 
7 
 

 

Esta mutación había sido asociada previamente con la AOA4 (Bras et al., 2015).  Por esta 

razón, es muy probable que en la población costarricense se encuentren familias con 

AOA4, que carguen la mutación Thr408del también en condición homocigota.  

El posible efecto de las mutaciones en el desarrollo de los fenotipos de enfermedad es 

uno de los focos de interés en la investigación en genética humana.  El desafío de 

determinar si las manifestaciones fenotípicas están asociadas con grupos específicos de 

residuos que comparten una relación subyacente representa una oportunidad para estudiar 

cómo surgen las diferentes manifestaciones fenotípicas. Sin embargo, los enfoques 

bioinformáticos recientes en el estudio de las proteínas y sus relaciones subyacentes han 

abierto una puerta para que en este proyecto se proponga llevar a cabo un análisis con las 

familias proteicas involucradas en los diferentes tipos de AOA. Esto mediante un enfoque 

computacional que buscará agrupar residuos proteicos e identificar aquellos sitios donde 

exista coevolución, esto se refiere a los cambios coordinados que pueden producirse en 

biomoléculas, por lo general para mantener o refinar las interacciones funcionales entre 

esos pares (De Juan, Pazos, & Valencia, 2013) y de esta manera establecer una asociación 

con los fenotipos clínicos reportados así como identificar nuevos residuos de interés que 

compartan patrones de coevolución con los residuos que sean asociados a fenotipos de la 

AOA.  

 

 

 

 

 



 
8 
 

 

Metodología 

ANÁLISIS MOLECULAR 

 
Comité de ética  

Esta investigación es parte del proyecto CMT que es dirigido por el Dr.rer.nat. Alejandro 

Leal Esquivel, el cual se encuentra inscrito ante la Vicerrectoría de Investigación y cuenta 

con aprobación del comité ético científico de la UCR.  

Selección de pacientes 

El proceso de reclutamiento de pacientes comenzó con la generación de una base de datos 

con información de contacto (correo electrónico, número de teléfono, hospital(es) en el 

que labora) de cada uno de los neurólogos registrados a nivel nacional en la Asociación 

Costarricense de Ciencias Neurológicas, además se cotejaron estos datos con los de los 

directorios de diferentes hospitales del ámbito público y privado en el país. Estos datos 

se utilizaron para establecer líneas de comunicación con todos los neurólogos, se les 

informó sobre el proyecto de investigación, y se les proporcionó el teléfono de la Escuela 

de Biología de la UCR, para que, ante un caso sospechoso de AOA, ellos le indicaran al 

paciente que si lo deseaban podían comunicarse con el equipo investigador.  

A cada voluntario se le asignó un código para asegurar la anonimidad de las muestras y 

cada expediente fue revisados para determinar quiénes contaban con características 

fenotípicas que se correspondían a las esperadas en ataxia. Una vez completado el estudio 

de casos, se realizaron dos viajes, uno a Pérez Zeledón y otro a la provincia de Alajuela 

para tomar muestras de saliva de los familiares de los pacientes seleccionados con el kit 

prepIT.L2P y se procedió a la fase de análisis de dichas muestras. 



 
9 
 

 

Adicional a este reclutamiento se solicitó a la Vicerrectoría de Investigación el acceso a 

muestras que se encuentran en el Instituto de Investigaciones en Salud (INISA) y que 

forman parte de otro estudio de neuropatías periféricas realizado en dicha Institución.  

 
 
Obtención y procesamiento de muestras 

El procesamiento comenzó con la extracción de ADN de las 48 muestras obtenidas 

durante el reclutamiento, para las muestras de sangre se utilizó el kit QIAamp DNA Blood 

Mini Kit (250) y para las muestras de saliva se utilizó el protocolo de purificación de 

ADN del kit prepIT.L2P. Una vez extraído el ADN de todas las muestras se procedió a 

la debida cuantificación de ADN de las mismas con el equipo Thermo Scientific 

NanoDrop 2000 Uv-vis Spectrophotometer. Todas las muestras contaron con una 

concentración de ADN adecuada para avanzar a la fase de pruebas de Reacción en Cadena 

de la Polimerasa  para PNKP. 

El siguiente paso fue la purificación de los productos de las amplificaciones siguiendo el 

protocolo disponible en el manual de procesos del laboratorio 240 de la Escuela de 

Biología de la UCR que se basa en un método orgánico a base de sales. Una vez realizada 

dicha purificación se procedió con el análisis de secuenciación bidireccional de Sanger 

mediante el método dye terminator, para el que se utilizó el secuenciador 3130 Genetic 

Analyzer (Applied Biosystems) que se encuentra en el laboratorio 240 y con  los 

iniciadores PNKP-del-F GGGTTTGTGTTGTCGATGG, PNKP-del-R 

TCTGCCGATCTGTTTGTGAC para la mutación Thr408del y PNKP-50364522-F 

ATGTCTAAAGTGCTCATGCCAGG y PNKP-50364522-R 

GGTACTGTTGGGGATAGCAGG para la mutación p.Gln517X..  

 

 

 



 
10 

 

 

Detección de variantes 

Todas las regiones de exones de todos los genes humanos (~22.000) fueron capturadas 

por xGen Exome Research Panel v2 (Integrated DNA Technologies, Coralville, Iowa, 

USA). Las regiones capturadas del genoma se secuenciaron con Novaseq 6000 (Illumina, 

San Diego, CA, EE.UU.). El análisis de los datos brutos de la secuenciación del genoma, 

incluida la alineación con el genoma humano de referencia GRCh37/hg19, la llamada de 

variantes y la anotación, se llevó a cabo con herramientas bioinformáticas de código 

abierto y software interno perteneciente a 3billion. El software de interpretación 

automática de variantes, EVIDENCE, se utilizó para priorizar las variantes basándose en 

las directrices del ACMG (Richards et al., 2015) y en el fenotipo de cada paciente. Este 

sistema tiene tres pasos principales; filtración de variantes, clasificación y puntuación de 

similitud para fenotipo del paciente (Seo et al., 2020). En primer lugar, gnomAD 

(Karczewski et al., 2020) como base de datos del genoma de la población y la base de 

datos de 3 mil millones de genomas se utilizaron para estimar la frecuencia alélica. 

Filtrado y validación 

Las variantes comunes con una frecuencia alélica menor de >5% se filtraron de acuerdo 

con la BA1 de la directriz del ACMG (Richards et al., 2015). En segundo lugar, se 

extrajeron datos de evidencia sobre la patogenicidad de variantes de una serie de 

literaturas científicas y bases de datos de enfermedades, incluyendo ClinVar (Landrum 

et al., 2018) y UniProt (Bateman et al., 2021). La patogenicidad de cada variante en sus 

enfermedades asociadas se evaluó de acuerdo con las recomendaciones de la directriz del 

ACMG (Richards et al., 2015). En tercer lugar, los fenotipos clínicos de los pacientes se 

transformaron en los correspondientes términos estandarizados de la ontología del 

fenotipo humano (Köhler et al., 2021) y se accedió a ellos para medir la similitud (Greene, 

Richardson, & Turro, 2016; Köhler et al., 2009) con cada una de las ~7.000 enfermedades 



 
11 

 

 

genéticas raras (Online Mendelian Inheritance in Man, OMIM®. McKusick-Nathans 

Institute of  Genetic Medicine, Johns Hopkins University (Baltimore, MD). La 

puntuación de similitud entre el fenotipo de cada paciente y los síntomas asociados a esa 

enfermedad, causados por variantes priorizadas usando como base aquellas asociadas a 

desórdenes neurológicos y según las directrices del ACMG, osciló entre 0 y 10. Las 

variantes de un solo nucleótido se confirmaron mediante secuenciación bidireccional de 

Sanger. 

 
Reacción en cadena de la polimerasa (PCR) de largo alcance 
 
En el caso del individuo AT-004, se realizó una PCR de largo alcance para la GAA-TR 

expandida en el gen FXN utilizando el kit Quantabio AccuStart Long Range SuperMix 

bajo las condiciones especificadas en el protocolo proporcionado dentro del mismo. 

Análisis clínicos 

Cinco individuos afectados por una neuropatía periférica no diagnosticada (AT-004, AT-

010, AT-026, AT-042 y 1071) se sometieron a una evaluación clínica por parte de 

neurólogo. Se realizó un examen clínico estándar, que incluía una revisión de la historia 

clínica completa de cada paciente y exámenes físicos para medir los reflejos, el tono 

muscular, los movimientos oculares, la afectación sensorial y otros signos que ayudaran 

a establecer una relación fenotipo-genotipo. 

 
ANÁLISIS BIOINFORMATICO 

 
Se realizó el análisis de información mutual para la identificación de residuos 

coevolucionados y correlacionados con otros residuos dentro de las mismas secuencias 

proteicas en las diferentes superfamilias a las que pertenecen las proteínas asociadas con 

AOA, siguiendo las especificaciones de Xia et al (2017).  

 
 
 



 
12 

 

 

Obtención de datos 

Usando el sitio de Centro Nacional para la Información en Biotecnología (Altschul et al., 

1997), se obtuvieron las secuencias de las superfamilias proteicas tomando como 

referencia las secuencias o números de acceso Q7Z2E3 (APTX), Q7Z333 (SETX), 

Q8WYR1 (PIK3R5) y Q96T60 (PNKP) del Homo sapiens , ya que el ser humano es el 

caso de estudio y el sistema biológico a analizar. Utilizando PSI-BLAST (Altschul et al., 

1997), con los parámetros predeterminados, se llevaron a cabo 3 iteraciones. El resultado 

fue un output compuesto de aproximadamente de 500 secuencias proteicas de cada 

superfamilia. La redundancia de las secuencias se redujo mediante CD-HIT (programa 

de agrupamiento por similitud) (Li, Jaroszewski, & Godzik, 2001) utilizando un corte de 

identidad de secuencia del 95% para mayor astringencia. El resultado obtenido fue un set 

de trabajo que contenía al menos 150 secuencias de cada superfamilia proteica.  

 

Curado de datos y alineamientos de secuencias 

Para disminuir las posibilidades de una sobrerrepresentación de proteínas similares con 

diferentes números de entrada se utilizó la herramienta MEGA7 (Kumar, Stecher, 

Tamura, & Dudley, 2016) y se realizó un curado manual de los datos para asegurar la 

generación de un conjunto de datos de secuencia de alta calidad para la alineación. 

Usando el software ClustalW (Thompson, Gibson, & Higgins, 2002) se alinearon cada 

uno de los conjuntos de datos de trabajo. Para curar aún más los archivos de los 

alineamientos se realizaron ajustes manuales de acuerdo a los datos obtenidos. 

Como último paso de esta fase se llevó a cabo un alineamiento de refinamiento por medio 

del software MUSCLE (Edgar, 2004). 

 

 

https://www.uniprot.org/uniprot/Q7Z2E3
https://www.uniprot.org/uniprot/Q7Z333
https://www.uniprot.org/uniprot/Q8WYR1
https://www.uniprot.org/uniprot/Q96T60


 
13 

 

 

Análisis de información mutual 

Mediante el uso de MISTIC (Servidor de Información Mutua para Inferir Coevolución) 

se determinó la correlación entre dos posiciones de residuos en los archivos de las 

múltiples secuencias. La identificación de los pares de residuos que tienen correlaciones 

evolutivas significativas (valores por encima del umbral de 6.5), fueron definidos por 

MISTIC.  

Los residuos se visualizaron por medio de un gráfico Circos y una red en Cytoscape 

(incorporados dentro de MISTIC), donde aquellos aminoácidos significativos se 

reportaron de acuerdo a una escala de colores que representa su conservación evolutiva 

y su correlación, esto de acuerdo a los descrito por (Simonetti, Teppa, Chernomoretz, 

Nielsen, & Marino Buslje, 2013). 

 

Análisis de predicción de estabilidad proteica 

Para predecir los posibles efectos causados en la estabilidad de los aminoácidos por los 

cambios de secuencia, utilizamos el iStable 2.0. Esta herramienta integra 11 predictores 

diferentes en un único sistema. 

Los modelos de iStable 2.0 se basan en clasificadores binarios (0/1) para la estabilidad o 

la inestabilidad y en un regresor para los valores continuos que utiliza características 

basadas en la secuencia para calcular la predicción de delta delta G (ddG), así como la 

estabilidad y la inestabilidad después de un cambio de aminoácido (Chen et al., 2020).  

Predicción de funcionalidad proteica 

Para predecir el potencial patogénico de las alteraciones de la secuencia de ADN 

encontradas en el paciente AT-042, se utilizó la herramienta de software MutationTaster 

(Schwarz, et al., 2014). Se analizó la variante presentada en la posición dada en todas las 



 
14 

 

 

transcripciones factibles de Ensembl (Howe et al., 2021), utilizando modelos Random 

Forest para las predicciones. El voto del árbol indica cuántos árboles de decisión del 

bosque aleatorio sugieren una alteración deletérea frente a cuántos sugieren una 

alteración benigna (Schwarz et al., 2014).  

 

 

 

 

 

 

 

 

 

 



 
15 

 

 

Objetivo General 

Determinar la causa genética de las AOAs en pacientes costarricenses por medio de 

secuenciación, así como relacionar el fenotipo de los pacientes con las características de 

conservación y correlación de residuos en las superfamilias de las proteínas asociadas 

con AOA mediante el uso de herramientas bioinformáticas, esto con el fin de contribuir 

con una clasificación clínica más certera de las AOAs en Costa Rica. 

Objetivos Específicos 

• Identificar mutaciones en genes involucrados en AOA en pacientes costarricenses 

mediante los métodos de secuenciación de Sanger y secuenciación de nueva generación, 

para generar datos sobre las mutaciones causantes de la AOAs en la población del país. 

• Establecer la relación fenotipo-genotipo en las familias con AOA en Costa Rica mediante 

análisis clínicos, con el fin de poder formular recomendaciones posteriores para un 

diagnóstico eficiente de la causa genética de la enfermedad. 

• Identificar residuos coevolutivos conservados en cada una de las proteínas involucradas 

en los diferentes tipos de AOA, así como las interacciones entre aminoácidos y residuos 

coevolutivos en las diferentes familias de proteínas en estudio, mediante el análisis de 

información mutual. Esto con el fin de establecer una asociación con los fenotipos 

clínicos, y de presentarse el caso, identificar nuevos residuos de interés que compartan 

patrones de coevolución con los residuos que sean asociados a fenotipos de las AOAs. 

Este trabajo sentaría las bases para llevar a cabo el diagnóstico molecular de las 

neuropatías y ataxias asociadas a este gen, así como en otros genes involucrados en AOA 

y generaría datos fenotípicos relacionados con AOA en Costa Rica. 



 
16 

 

 

Artículo:  

ANALYSIS OF ATAXIA IN COSTA RICA THROUGH AN INTEGRATIVE AP-
PROACH 

 
Hilda Torres-Ulate1, Sixto Bogantes2, Eugene Lee3, Go Hun Seo3, José Guevara-Coto4, 

Gabriela Chavarría5, Alejandro Leal5 

 

1 School of Biology, University of Costa Rica. 11501 San José, Costa Rica. 
2 Faculty of Medicine, University of Costa Rica. 11501 San José, Costa Rica. 

3 3billion, Inc. 06193, Seoul, Republic of Korea 
4 School of Computer Science and Informatics, University of Costa Rica. 11501 San José, Costa 

Rica. 
5 Section of Genetics and Biotechnology, School of Biology, University of Costa Rica. 11501 

San José, Costa Rica. 
 

Abstract 

The ataxias are a group of clinically heterogeneous disorders, which may or may not be 
heritable depending on the etiology. The clinical manifestations of the hereditary ataxias 
are progressive incoordination of movement and speech (dysarthria), and an unsteady, 
uncoordinated, broad-based gait. In addition, patients may develop ophthalmoplegia (eye 
movement limitations), spasticity, neuropathy, and cognitive difficulties. This paper 
focuses on ataxia with oculomotor apraxia (AOA), which is an autosomal recessively 
inherited disease. Four types of AOA (AOA1-AOA4) have been described to date. Since 
information on AOA-causing mutations in Costa Rica is scarce, here we characterize 
AOA-causing mutations in Costa Rican patients using Sanger sequencing and next-
generation sequencing methods. The challenge of determining whether phenotypic 
manifestations are associated with specific groups of residues that share an underlying 
relationship has not yet been validated. To address this, we constructed a multiple 
sequence alignment and analyzed the resulting files with mutual information. In each 
protein family, Mutual Information (MI) identified pairs of residues with functional 
importance due to their location within important domains and regions. This work would 
contribute to characterizing the prevalent mutations in patients suffering from ataxia with 
oculomotor apraxia in the country, as well as the associated phenotype, and would lay 
the groundwork for developing working protocols for the molecular diagnosis of the 
disease in Costa Rica. 

Keywords 

Ataxia, apraxia, dysarthria, peripheral neuropathy, molecular genetics, sequencing, 
multiple sequence alignment, mutual information, coevolution. 

 
 
Introduction 
 
Ataxias are a group of disorders within cerebellar diseases. The first efforts to classify 
them were made by  Gordon Holmes (1908) back in the mid-nineteenth and early 
twentieth century . These can be clinically and genetically heterogeneous, with slowly 



 
17 

 

 

progressive gait incoordination as its main trait. Ataxias can be non-hereditary, when they 
appear in adulthood (Klockgether, 2010), and are caused by external factors such as 
chronic alcoholism, vitamin deficiencies, vascular disease, primary or metastatic tumors 
and paraneoplastic diseases associated with occult carcinoma of the ovary, breast or lung, 
idiopathic degenerative disease and multiple system atrophy (spinal muscular atrophy) 
(Jayadev & Bird, 2013). Ataxias can also be hereditary when associated with a causal 
pathogenic variation in certain genes. These can cause, among other things, cerebellar 
atrophy, gait and speech problems of varying severity.  In addition, individuals may 
develop eye movement limitations (ophthalmoplegia), spasticity, neuropathy, and 
cognitive difficulties (Jayadev & Bird, 2013).  
 
For inherited disorders, all three modes of inheritance can be observed (Perlman, 2022). 
Regarding the autosomal dominant ataxias, there are more than 30 forms known to this 
date, with a prevalence of 1.2 to 1.9 per hundred thousand inhabitants in countries such 
as Spain and Brazil, respectively (Ruano et al., 2014).  Showing this pattern of 
inheritance, spinocerebellar ataxias (SCA) are a group of progressive ataxia disorders 
caused by degeneration of the cerebellum and its afferent and efferent connections 
(Schöls et al; 2004),  linked to a CAG triplet repeat expansion mutation in coding regions 
of genes (Manto, 2005; Pulst et al., 1996). The prevalence of SCA is estimated to be 
approximately 1-5 per 100,000 inhabitants (Ruano, et al., 2014). The most common are 
SCA1, SCA2, SCA3 y SCA6 (Schmitz-Hübsch et al., 2008) with SCA3 being the most 
common worldwide (Bird, 2019).   This may vary in certain regions due to a founder 
effect, as exemplified by SCA2 in Cuba and SCA10 in Mexico (Manto, 2005).  
 
There are large variations among SCA subtypes (Schöls, Ludger; Bauer, Peter; Schmidt, 
Thorsten; Schulte Thorsten; Riess, 2004), symptoms of SCA1, SCA2, SCA3, SCA7, 
SCA8, SCA12, SCA13, SCA17, or SCA25 may begin in the first decade, whereas ataxia 
may appear after the age of 65 years in SCA6. Regarding SCA17 the penetrance is 
reduced (Zühlke et al., 2003) while SCA8 shows a complex inheritance pattern with 
extremes of incomplete penetrance, often with only one or two affected individuals in a 
family (Manto, 2005). 
 
Symptoms commonly developed by individuals with these conditions are a slowly 
progressive cerebellar syndrome with various combinations of oculomotor disorders, 
dysarthria, dysmetria/kinetic tremor and/or ataxic gait. They may also be affected with 
pigmentary retinopathy, extrapyramidal movement disorders (parkinsonism, dyskinesias, 
dystonia, chorea), pyramidal signs, cortical symptoms (seizures, cognitive 
impairment/behavioral symptoms) and peripheral neuropathy (Table 1). An overlap of 
phenotypes between the different subtypes can occur due to SCAs being genetically 
heterogeneous. (Manto, 2005). 
 
Regarding ataxias of autosomal recessive inheritance, these account for approximately 3 
cases per 100,000 inhabitants, with more than 60 forms being Friedreich's ataxia, ataxia-
telangiectasia and oculomotor apraxia being the most common of them (Ruano et al., 
2014). Recessive ataxias have a reported prevalence in Spain of 7.2 per 100,000 
inhabitants (Ruano et al., 2014). Many of the hereditary ataxias have overlapping 
presentations and there is a high degree of genetic heterogeneity (Inlora et al., 2017).  
 



 
18 

 

 

Table 1. Clinical presentation of the most common SCAs 

Signs and symptoms SCA
1 

SCA2 SCA3 SCA
4 

SCA7 SCA8 SCA10 SCA12 SCA1
3 

SCA17 SCA18 SCA2
5 

Oculomotor disorder X X X  X        

Dysmetria/kinetic tremor      X       

Pigmentary retinopathy      X        

Extrapyramidal movement 
disorders 

X X X     X  X   

Pyramidal signs X X X X X X  X X    

Peripheral neuropathy X X X X  X  X   X X 

Cognitive disability X X X      X X   

Seizures       X   X   

Source: Schöls, Ludger; Bauer, Peter; Schmidt, Thorsten; Schulte Thorsten; Riess, 2004. 
 
 
Within the autosomal recessively inherited disorders, during the last fifteen years the 
ataxia with oculomotor apraxia (AOA) has been described, this is a group involving 
axonal sensorimotor neuropathy, and extrapyramidal features (Bras et al., 2015). These 
were identified as the second most prevalent collection of autosomal recessive ataxias 
worldwide (Ruano et al., 2014) with an incidence varying depending on the type of AOA. 
They are caused by mutations in genes responsible for DNA repair, premature protein 
termination, protein maturation deficiencies, or both. Four types have been described to 
date: AOA1, AOA2, AOA3 (Tassan et al., 2012) and AOA4/CMT2B2 (Leal et al., 2018). 
The different types of AOA share many symptoms with each other (Inlora et al., 2017) 
(Table 2).  
 
AOA1 has an age of onset ranging from infancy to preadolescence (2-12 years) and is 
caused by mutations in the Aprataxin gen (APTX) (Coutinho & Barbot, 2002), which is 
located on 9p21.1; it has 7 exons and a size of 28,217 bp. (Kent et al., 2002).  APTX 
encodes for the protein aprataxin (APTX), which is a nuclear protein, present in both the 
nucleoplasm and nucleolus. It is responsible for catalyzing the nucleophilic release of 
adenylate groups covalently attached to the 5' phosphate ends at single-strand breaks, 
resulting in the production of 5' phosphate ends that can be efficiently joined. It is 
associated with the other DNA repair proteins, playing a role in single-stranded DNA 
repair through its nucleotide-binding activity and its diadenosine polyphosphate 
hydrolase activity  (MIM 2089200; Prasad et al., 2009). 
 
Németh et al. (2000) first  identified a novel locus for a primary autosomal recessive 
cerebellar ataxia in the SETX gene know as AOA2, with an average age of onset in pre-
adolescence (12.7 years). (Schiess et al., 2017). SETX is located at 9q34.13, has 26 exons 
and is 93,630 bp in size (Bateman et al., 2021). It encodes for a protein that, because of 
its homology to the fungal Sen1p protein, was named Senataxin. It has RNA helicase 
activity encoded by a DNA/RNA helicase domain at the C-terminal end, suggesting that 



 
19 

 

 

it may be involved in DNA and RNA processing (Prasad et al., 2009). In association with 
Rrp45, directs the RNA exosome complex to sites of transcription-induced DNA damage 
(Richard, Feng, & Manley, 2013b). Contains at its C-terminal end a classic seven-motif 
domain found in superfamily 1 of the helicase (M. C. Moreira et al., 2004). This domain 
has strong homology with the human RENT1 and IGHMBP2 genes, which are two genes 
encoding proteins known to have functions in RNA processing (Becherel et al., 2015).    
 

Table 2. Signs presented by the different types of AOA 

Sign AOA type 

AOA1 AOA2 AOA3 AOA4/CMT2B2 

Evolution Severe Benign Benign Severe 

Oculomotor apraxia X X X X 

Dystonia X X Not mentioned X 

Axonal neuropathy X X X X 

Cognitive disability X Not mentioned Not mentioned X 

Cerebellar atrophy X X X X 

Chorea X X Not mentioned Not mentioned 

Dysmetria Not mentioned Not mentioned X Not mentioned 

Dysarthria X X X X 

Saccadic eye movements X X X X 

Source: Coutinho & Barbot, 2002; Tassan et al., 2012; Bras et al., 2015; Szpisjak, Obal, En-
gelhardt, Vecsei, & Klivenyi, 2016; Inlora et al., 2017; Mariani et al., 2017; Schiess et al., 2017 
 
Tassan et al. (2012) identified AOA3 in a consanguineous family in Saudi Arabia, whose 
affected members had clinical features similar to those of individuals with AOA2, but 
with a mean age of onset in adolescence (15.6 years). This type of ataxia is caused by 
mutations in PIK3R5, which is located at 17p13.1.  It has 18 exons and a size of 30,869 
bp (Kent et al., 2002). This gene encodes regulatory subunit 5 of the phosphatidylinositol 
3-kinase gamma class complex (PIK3γ), which is a 101 kD regulatory subunit of the class 
I PIK3γ complex. PIK3R5 is a dimeric enzyme, consisting of a 110 kD catalytic gamma 
subunit and a 55.87 or 101 kD regulatory subunit that interacts with class 1B 
phosphoinositide-3-kinase (PI3K) (Uhlen et al., 2015). It acts through high-affinity 
interaction with G-beta-gamma proteins to recruit the catalytic subunit from the cytosol 
to the plasma membrane (Uhlen et al., 2015).  Phosphatidylinositol 3-kinase (PI3K) is a 
member of a unique and highly conserved family of intracellular kinases that 
phosphorylate the 3'-hydroxyl group of phosphatidylinositol and phosphoinositides. This 
reaction results in the activation of multiple intracellular signaling pathways that regulate 
diverse and important functions, such as cellular metabolism, survival and polarity, and 
vesicle trafficking (Engelman et al., 2006).  
 
Within the rare group of cerebellar ataxias to which AOA belongs, patients with AOA 
associated to mutations in PNKP have an average age of disease onset in childhood (4.3 
years) (Schiess et al., 2017) and in some cases is characterized by cerebellar ataxia, 
oculomotor apraxia, polyneuropathy, and cerebellar atrophy at MRI. Such phenotype was 



 
20 

 

 

classified as ataxia with oculomotor apraxia type 4 (AOA4) (Gatti et al., 2019). In other 
cases, the presenting clinical phenotype was characterized by the presence of a motor and 
sensory axonal polyneuropathy, resembling a form of Charcot–Marie–Tooth disease and 
identified as CMT2B2 (Leal et al., 2018; Pedroso et al., 2015). In these patients, slurred 
speech and cerebellar atrophy at brain MRI were also described (Leal et al., 2018).  
 
PNKP is located at 19q13.33, has 16 exons and is 6,337 bp in size and encodes for a 
polynucleotide kinase 3' phosphatase, which in response to ionizing radiation or oxidative 
damage, catalyzes the 5' phosphorylation of nucleic acids and also has an associated 3' 
phosphatase activity, which predicts an important role in DNA repair after ionizing 
radiation or oxidative damage (Gatti et al., 2019; Leal et al., 2018; Uhlen et al., 2015). 
 
In view of the foregoing on the different phenotypes that can be identified in the same 
disorder, to establish the diagnosis of hereditary ataxia, the following are recommended: 
detecting typical clinical signs on neurological examination, including poorly 
coordinated hand movements, dysarthria, eye movement abnormalities, identification of 
distinctive features on magnetic resonance imaging or computed tomography scans 
(Klockgether, 2010; Perlman, 2022), documentation of the possible hereditary nature of 
the disease by developing a family history searching for cases of relatives with ataxia, 
identifying a mutation causing ataxia or recognizing a clinical phenotype characteristic 
of a genetic form of ataxia (Jayadev & Bird, 2013; Perlman, 2022). If a genetic cause 
cannot be identified, it is necessary to proceed with the exclusion of non-genetic causes 
of ataxia by means of CT scans, magnetic resonance imaging and detection of multiple 
antibodies (Klockgether, 2010).   
 
Since the challenge of determining whether phenotypic manifestations are associated 
with specific groups of residues that share an underlying relationship, the possible effect 
of mutations in the development of disease phenotypes is a focus of interest and 
represents an opportunity to study how different phenotypic manifestations arise. 
 
As a bioinformatics approach in the study of proteins and their underlying relationships, 
mutual information can be used to estimate the extent of the mutual coevolutionary 
relationship between two positions in a protein family and it can be applied to predict 
positional correlations in a multiple sequence alignment to make possible the analysis of 
those positions structurally or functionally important in a given fold or protein family 
(Dunn, Wahl, Gloor; 2008), such as the ones associated with Ataxia with Oculomotor 
Apraxia (AOA). An integrative approach based of analyzing full-length sequence 
alignments from proteins and the subsequent identification of protein sectors using 
statistical analysis has been proposed in previous studies (Guevara Coto, Schwartz, & 
Wang, 2014; Halabi et al, 2009; McLaughlin et al, 2012; Morcos et al., 2011) aiming to 
cluster amino acid residues based on an underlying interaction, such as co-evolution, to 
understand how mutations in these related sites can lead to different disease phenotypes.  
 
Given that AOAs have not been studied in Costa Rica this project aimed to molecularly 
characterize the variants associated with ataxia in the population of the country, as well 
as to study the relationship between these variants and the protein conformation they 
produce, with the phenotype of patients.   

 



 
21 

 

 

Methodology 
 
Ethics Committee  
This research is part of the Project “Determination of genes related to peripheral 
neuropathies and a group of ataxias”  (No. 111-B8-37), registered with the Vice Rector's 
Office for Research and has been approved by the Ethics Committee.  
 
Recruitment process 
Neurologists registered in the main hospitals of the country were contacted and informed 
about the project so that, in the event of a suspected case of AOA, they could indicate to 
the patient that if they wished, they could contact the research team.  
 
In addition to this recruitment, the Ethics Committee was asked form permission to access 
to samples from the Institute of Health Research (INISA) of the University of Costa Rica, 
which were part of another study on ataxias carried out in that institution.  
 
Neurologists working in Costa Rica, invited patients to establishing contact with our 
research group.  After receiving a detailed explanation about the project, if the individuals 
agreed to participate in the study, they proceeded to sign an informed consent form and 
a code was assigned to ensure the anonymity of the samples. Individuals with ataxia or 
probable ataxia phenotypic were taken into account.  Saliva or blood samples were 
collected. 
 
Variants detection 
 
We obtained DNA samples of 48 individuals who showed phenotypic characteristics that 
corresponded to those expected in ataxia and did not have a definitive clinical diagnosis 
provided by a neurologist. Of these, 28 samples met the quality criteria established to 
allow them to continue to be processed, so all exon regions of all human genes were 
captured by xGen Exome Research Panel v2 (Integrated DNA Technologies, Coralville, 
Iowa, USA). The captured regions of the genome were sequenced with Novaseq 6000 
(Illumina, San Diego, CA, USA). Analysis of the raw genome sequencing data, including 
alignment with the GRCh37/hg19 human reference genome, variant calling and 
annotation, was performed with open-source bioinformatics tools and 3Billion in-house 
software. The automatic variant interpretation software, EVIDENCE has three main 
steps; variant filtering, ranking, and similarity scoring for patient phenotype (Seo et al., 
2020). First, gnomAD (Karczewski et al., 2020) as a population genome database and the 
3Billion genome database were used to estimate allele frequency. 
 
Filtering and Validation 
 
Common variants with an allele frequency less than >5% were filtered according to BA1 
of the ACMG guideline (Richards et al., 2015). Second, evidence data on variant 
pathogenicity were extracted from a range of scientific literatures and disease databases, 
including ClinVar (Landrum et al., 2018) and UniProt (Bateman et al., 2021). The 
pathogenicity of each variant in its associated diseases was assessed according to the 
recommendations of the ACMG guideline (Richards et al., 2015). Third, patients' clinical 
phenotypes were transformed into corresponding standardized human phenotype 
ontology terms (Köhler et al., 2021) and accessed to measure similarity (Greene et al., 
2016; Köhler et al., 2009) to each of the ~7,000 rare genetic diseases rare genetic diseases 



 
22 

 

 

(Online Mendelian Inheritance in Man, OMIM®. McKusick-Nathans Institute of Genetic 
Medicine, Johns Hopkins University (Baltimore, MD). The similarity score between each 
patient's phenotype and the symptoms associated with that disease caused by the 
prioritized variants using as a starting point those associated with neurological diseases 
and according to ACMG guidelines ranged from 0 to 10. The single nucleotide variants 
and all indels were confirmed by bidirectional Sanger sequencing. 
 
Long range polymerase chain reaction (PCR) 
 
In the case of individual AT-004, a long PCR for expanded GAA-TR in the FXN gene 
was performed using Quantabio AccuStart Long Range SuperMix Kit under the 
conditions specified in the protocol provided in the package. 
 
Clinical analysis 
 
Because some samples belonged to a previous study, these individuals could not be 
located for clinical evaluation, thus only five individuals affected with an undiagnosed 
peripheral neuropathy (AT-004, AT-010, AT-026, AT-042 and 1071) underwent clinical 
analysis. A standard clinical examination was performed, which included a review of the 
complete clinical history of each patient and physical examinations to measure reflexes, 
muscle tone, eye movements, sensory involvement and other signs that would help to 
establish a phenotype-genotype relationship. 
 
Data Acquisition 
 
For the proteins associated to AOA, the sequences of the protein superfamilies were 
obtained from the National Center for Biotechnology Information (Altschul et al., 1997) 
using as reference the sequences or accession numbers Q7Z2E3 (APTX), Q7Z333 
(SETX), Q8WYR1 (PIK3R5) and Q96T60 (PNKP) from Homo sapiens, since the human 
being is the case study and the biological system to be analyzed. Using PSI-BLAST 
(Altschul et al., 1997), with the predetermined parameters, 3 iterations were carried out. 
The result was an output composed of approximately 500 protein sequences from each 
superfamily. The redundancy of the sequences was reduced by CD-HIT (similarity 
clustering program). (Li et al., 2001) using a 95% sequence identity cutoff for 
astringency. The result obtained was a working set containing at least 150 sequences from 
each protein superfamily. 
 
Data curation and sequence alignments 
 
To reduce the chances of overrepresentation of similar proteins with different entry 
numbers, the MEGA7 tool was used. (Kumar et al., 2016) and manual data curation was 
performed to ensure the generation of a high quality sequence dataset for alignment.  
Using ClustalW (Thompson et al., 2002) each of the working data sets were aligned. To 
further curate the alignment files, manual adjustments were made according to the data 
obtained.  As the last step of this phase, a refinement alignment was carried out using 
MUSCLE (Edgar, 2004). 
 
 
 
 

https://www.uniprot.org/uniprot/Q7Z2E3
https://www.uniprot.org/uniprot/Q7Z333
https://www.uniprot.org/uniprot/Q8WYR1
https://www.uniprot.org/uniprot/Q96T60


 
23 

 

 

Mutual information 
 
Using Mutual Information Server to Infer Coevolution (MISTIC), the correlation 
between two residue positions in the multiple sequence files was determined. The 
identification of pairs of residues that have significant evolutionary correlations (values 
above the threshold of 6.5) were defined by MISTIC.  Residues were visualized by means 
of a Circos plot and a network in Cytoscape (incorporated within MISTIC), where those 
significant amino acids were reported according to a color scale representing their 
evolutionary conservation and correlation, this according to those described by 
(Simonetti et al., 2013). 
 
Protein stability prediction analysis 
 
To predict the possible effects caused in amino acid stability caused by sequence changes, 
we used the iStable 2.0. The tool integrates 11 different predictors in a single system. The 
iStable 2.0 models are based on binary classifiers (0/1) for stability or instability and a 
regressor for continuous values that use sequence based features to calculate the to predict 
delta delta G (ddG) as well as stability and instability after an amino acid change (Chen 
et al., 2020). (Fig 1). 
 
Functionality prediction analysis 
 
To predict the pathogenic potential of DNA sequence alterations found in patient AT-
042, the software tool MutationTaster (Schwarz, et al., 2014) was used. It analyzed the 
submitted variant at the given position in all feasible Ensembl transcripts (Howe et al., 
2021), using Random Forest models for predictions. Tree vote indicates how many 
decision trees of the Random Forest are suggestive of deleteriousness vs. how many are 
suggestive of a benign alteration (Schwarz et al., 2014). 

 

 
 
 
 

Figure 1. System architecture with the key points of the process for protein stability prediction.  



 
24 

 

 

Results 

Phenotypic characterization  
 
All the examined patients have as a common factor the instability (ataxia), although they 
can be differentiated into two groups (Table 3): 
The first group includes individuals AT-004 and AT-026 where the ataxia is the most 
disabling clinical sign, with mild or absent signs of motor neuropathy, although they have 
phenotypically distinct forms of ataxia. AT-004 presents a sensory ataxia (neurophysiol-
ogy consistent with this modality) associated with pyramidal signs, without cerebellar 
involvement in neuroimaging and with non-motor manifestations of her disease suggest-
ing that it is a systemic condition where the cause of the ataxia is not a primary cerebellar 
disease. On the other hand, individual AT-026 presents a form of late-onset ataxia with 
pancerebellar symptoms and very important atrophic changes in his cerebellum evi-
denced in the neuroimaging suggesting a primary cerebellar disease as the cause of ataxia. 
 
The second group includes the other three individuals who are the most disabled ones, 
and where the functional limitation that has led to the loss of walking is associated with 
a clear peripheral motor nerve disease. All of them have lost walking autonomy and have 
trophic changes in the lower extremities (distal muscle atrophy). These individuals pre-
sent extrapyramidal signs of variable severity (in AT-042 and AT-010 are evident while 
in 1071 are mild). The signs of cerebellar dysfunction are mild. In this group AT-042 has 
a family history of three uncles suffering from a neurodegenerative disease with neurop-
athy and ataxia, one of them died at the age of 24. Otherwise, AT-010 was first diagnosed 
with cerebral palsy since he presented perinatal insult (hypoglycemia), his severe disa-
bility can be explained by this reason. Individual 1071 has normal cognitive function with 
a university education but is currently retired due to the severity of peripheral motor in-
volvement, she also has subtle signs of brainstem dysfunction. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 



 
25 

 

 

Table 3. Clinical evaluation result for the evaluated patients. 

 Group 1 Group 2 
 (AT-004) (AT- 026) AT-010 (1071) (AT- 042) 
Gender Female Male Male Female Male 
Age 22 70 26 44 21 
Decade of age at 
onset 

First Third First Third First 

First sign Syncope Polyneuropathy Polyneuropathy Polyneuropathy Choreathetosis 

More prominent 
sign 

Ataxia Polyneuropathy Polyneuropathy Polyneuropathy Polyneuropathy 

OMA - + ++ +/- - 
Slurred speech + + +++ + + 
Mobility Mobile, severe gait 

ataxia 
Mobile, mild gait 

ataxia 
Wheelchair Wheelchair Wheelchair 

Dystonia - - + - ++ 
Cognitive 
impairment 

- - Mild - Mild 

UE Muscle 
strength (I/D) 

5/5 4/3 4/1 3/0 5/5 

LE Muscle 
strength (P/I/D) 

5/5/5 5/3/0 1/0/0 5/0/0 4/0/0 

Reflexes 
(UE/Knee/Ankl
e) 

2/0/0 2/3/3 2/0/0 1/0/0 2/0/0 

UE Sensory 
involvement 
(T/Pa/V/Po) 

1/1/1/1 2/1/0/0 0/0/0/0 2/1/0/0 2/2/1/1 

LE Sensory 
involvement 
(T/Pa/V/Po) 

1/1/1/0 1/0/0/0 0/0/0/0 0/0/0/0 1/1/1/1 

Atrophy None Intrinsic muscles of 
hands and feet, 

calves 

Anterior and 
posterior 

compartments of 
forearm, intrinsic 
muscles of hands 
and feet, calves 

Intrinsic muscles of 
hands and feet, 

calves 

Calves 

Others Marcus Gunn sign 
(relative afferent 
pupillary defect) 

REM sleep behavior 
disorder 

Lymphedema Lymphedema, right 
hemifacial spams 

Three uncles affected 
with clinically similar 

disease 

Deformities Left ventricular 
hypertrophy  

Claw hand, pes 
cavus and 

hammertoes 

Claw and dropped 
hand, scoliosis 

Claw hand Head hyperextension, 
scoliosis 

Pyramidal signs Babinski - - - - 
Obesity ++ - - - - 
MRI findings Normal CA, no WM 

abnormalities 
CA, no WM 
abnormalities 

CA, no WM 
abnormalities 

CA, periventricular 
gliosis 

Electrophysiolo
gical data 

Axonal sensory 
neuropathy 

Normal - - Severe sensorimotor 
neuropathy 

OMA, oculomotor apraxia. UE, upper extremity; LE, lower extremity; P, proximal (knee extensor or flexor); I, intermediate 
(hand extensor or flexor (UE), foot extensor or flexor (LE)); D, distal (intrinsic hand muscles (UE), intrinsic foot muscles (LE)), 
Motor scale: 5, normal; 4, mild weakness; 3, ability to lift against gravity; 2, not able to lift against gravity; but movement visible, 
1, no movement, but tendon contraction visible; 0, complete paralysis, Reflexes/sense of vibration or position: 2, normal; 1, 
reduced; 0, absent; T, touch; Pa, pain; V, vibration; Po, position, Sensory involvement: 2, normal; 1, mildly reduced: distally to 
wrist level (UE) or malleoli level (LE), 0, severely reduced: distally to elbow level (UE) or knee level (LE). CA, cerebellar atrophy. 
WM, white matter. 
 
 

 
 
 
 



 
26 

 

 

Molecular characterization 
 
Five variants associated with peripheral neuropathies causing disorders were identified 
in 9 patients (Table 4).  

Table 4. Identified variants associated with peripheral neuropathies 

Sample Gen Variant Presentation Disorder 
AT-026 KCNC3 19-50826951-C-T 

NM_004977.3:c.1259G>A 
(NP_004968.2:p.Arg420His) 

Heterozygous Spinocerebellar 
ataxia 13 

AT-038 PMPCA 9-139313299-G-A 
NM_015160.2:c.1129G>A 

(NP_055975.1:p.Ala377Thr) 

Homozygous Spinocerebellar 
ataxia, autosomal 
recessive 2 
(SCAR2) 

AT-004 FXN Two expanded alleles of 908 and 
1116 triplets in intron 1 

Homozygous Friedreich ataxia 

AT-021 PNKP 19-50364522-G-A 
NM_007254.3:c.1549C>T 

(NP_009185.2:p.Gln517Ter) 

Homozygous AOA4/CMT2B 

1207 PNKP 19-50364522-G-A 
NM_007254.3:c.1549C>T 

(NP_009185.2:p.Gln517Ter) 

Homozygous AOA4/CMT2B 

AT-010 PNKP 19-50365103-CGTG-C 
NM_007254.3:c.1221_1223del 

(NP_009185.2:p.Thr408del) 

Homozygous AOA4 

AT-040 PNKP 19-50365103-CGTG-C 
NM_007254.3:c.1221_1223del 

(NP_009185.2:p.Thr408del) 

Homozygous AOA4/CMT2B 

AT-046 PNKP 19-50365103-CGTG-C 
NM_007254.3:c.1221_1223del 

(NP_009185.2:p.Thr408del) 

Homozygous AOA4/CMT2B 

1071 PNKP 19-50365103-CGTG-C 
NM_007254.3:c.1221_1223del 

(NP_009185.2:p.Thr408del)                                                                                                                          
19-50364522-G-A 

NM_007254.3:c.1549C>T 
(NP_009185.2:p.Gln517Ter) 

Compound 
Heterozygous 

CMT2B 

AT-042 PNKP 19-50365103-CGTG-C 
NM_007254.3:c.1221_1223del 

(NP_009185.2:p.Thr408del) 
19-50365031-G-GGC 

NM_007254.4:c.1294_1295dup 
(NP_009185.2:p.Arg433ProfsTer35) 

Compound 
Heterozygous 

AOA4 

 
The variant linked to Spinocerebellar Ataxia 13 (SCA13) in both a French and a Filipino 
families (M. F. Waters et al., 2005) was identified in one patient (AT-026): a transition 
in exon 2 c.G1259A  in the potassium channel, voltage-gated, shaw-related subfamily, 
member 3 gen (KCNC3) resulting in a substitution of an amino acids that is 100% 
conserved among members of the human KCNC family (M. F. Waters et al., 2006).  
KCNC3 is located within the linkage interval on chromosome 19q13.33, the voltage-
gated potassium channel plays an important role in the rapid repolarization of fast-firing 
brain neurons. The channel displays rapid activation and inactivation kinetics. It is 
involved in the regulation of the frequency, shape and duration of action potentials in 



 
27 

 

 

Purkinje cells and is required for normal survival of cerebellar neurons, therefore it is 
necessary for normal motor function (Bateman et al., 2021). 
 
Another variant identified in a patient (AT-038) is p.Ala377Thr, it was associated to 
Spinocerebellar ataxia, autosomal recessive 2 (SCAR2) in three families of Christian 
Lebanese Maronite origin (Jobling et al., 2015). It is caused by a missense mutation in 
exon 10, c.G1129A in the mitochondrial processing peptidase-alpha gene (PMPCA), 
leading to defective substrate recognition and binding leading to inadequate enzyme 
activity within the mitochondria (Jobling et al., 2015).  PMPCA cytogenetic location is 
9q34.3 (Bateman et al., 2021). It encodes the alpha subunit of the mitochondrial 
processing peptidase (MPP), which cleaves the targeting peptide of nuclear-encoded 
mitochondrial precursor proteins upon their import into mitochondria (Choquet et al., 
2016) 
 
Lastly, for seven patients, three variants were identified in PNKP, that is also located on 
chromosome 19q13.33 and contains a kinase and a phosphates domain that are both 
involved in DNA binding (Bateman et al., 2021) through its two catalytic activities which 
ensure that DNA termini are compatible with extension and ligation by either removing 
3'-phosphates from, or by phosphorylating 5'-hydroxyl groups on, the ribose sugar of the 
DNA backbone (Bateman et al., 2021). Two of those variants were already associated 
with AOA4 (Gatti et al., 2019).  
 
A transition c.C1549T in exon 17 of the polynucleotide kinase 3'-phosphatase (PNKP) 
gene. This variant causes a nonsense mutation (p.Gln517ter) predicted to truncate the last 
five amino acids (Leal et al., 2018). The second variant identified consists of a three-base 
deletion in exon 14, c.1221_1223del, this variant results in the deletion of residue Thr408 
(Thr408del) (MIM 616267). This variant was found in a homozygous presentation in 
three patients. 
 
Two patients were found to be compound heterozygous for variants in the PNKP gene. 
The first one presented the two variants already described in Costa Rica, p.Gln517ter and 
Thr408del. For the second patient, the variants found were Thr408del and Arg433Ter, a 
variant that has not yet been described and is predicted to cause a premature termination 
in the kinase region of the protein. It represents a nonsense mutation in exon 14, 
c.1294_1295dup. (Arg433ProfsTer35).   
 
 
Mutual information 
Mutual information (MI) identified functionally important positions known to cause 
AOA in three of the superfamilies. For APTX three positions were linked, for PIK3R5 
one position was recognized and for PNKP two position were identified.  In the case of 
SETX, six positions linked to genetically related disorders with AOA2 were found 
(Moreira & Koenig, 2018), four positions known to cause spinocerebellar ataxia with 
axonal neuropathy (SCAN2) and one position associated with Amyotrophic lateral 
sclerosis 4 (ALS4) respectively were identified as well (Bateman et al., 2021)(See table 
5). 
 



 
28 

 

 

Table 1. Functional important positions known to cause AOA in each superfamily. 

APTX SETX PIK3R5 PNKP/CMT2B 

206P, 247L and 
279W 

305W, 413P, 496P 
and 2213P (SCAN2) 

1554C (ALS4) 

629P 375G, 399L, 409C, 
424T, 429A, 439R, 
442G, 462R, 465E,  

515Y and 517Q 

 
Additionally, potential candidate positions that could give rise to disease phenotypes if 
mutated were found in all the super families (Table 2). 
  

Table 2.  Potential candidates that could cause AOA if mutated for each protein 
superfamily. 

 
APTX SETX PIK3R5 PNKP/CMT2B 

254P, 260H, 
262H, 319C 
and 322C 

4C, 6W, 7C, 38C, 40C, 43C, 55P, 62W, 
113P, 121P, 133C, 138C, 154P, 163P, 

170W, 195C, 208P, 225P, 238W, 242C, 
285P, 287W, 288P, 292C, 305W, 311P, 
342P, 352C, 370W, 375C, 376P, 379C, 
380P, 410W, 451C, 474C, 476H, 479W, 
485W, 492C, 524H, 535Y, 546G, 550G, 
555C, 568G, 573G, 574W, 587C, 612C, 

788C, 976F, 977P, 1093W 1266P, 
1270P, 1271P, 1278P, 1280P, 1293P, 
1318G, 1345S, 1497F, 1503P, 1509C, 
1549C, 1554C, 1565C, 1568H, 1594P, 
1599F, 1622P, 1708W, 1720G, 1721P, 
1722P, 1734P, 1737F, 1749P, 1763W, 
1804Y, 1805P, 1877C, 1909P, 1914F, 
1915C, 1916T, 1953P, 1959C, 1962H, 
1963G, 1964P, 2038C, 2039, 2047G, 

2067H, 2103G, 2152C, 2153C, 2159G, 
2160G, 2174P, 2177C, 2187C, 2194P, 
2199C, 2208P, 2212P, 2250C, 2261H, 
2262P, 2265C, 2267F, 2268P, 2288C, 
2292W, 2293P, 2294F, 2296P, 2384C, 
2389C, 2434W, 2476P, 2481P, 2491P 

and 2578P 

8C, 33W, 42W, 80P, 
115W, 116P, 118P, 120C, 
137P, 284P, 293W, 504P, 
545R, 546P, 560P, 616P, 
617W 634C, 663C, 726P, 

744W, 750W, 760C, 
818C, 833C, 838C, 865C, 

869C and 871P 

11W, 18G, 24L, 25P, 27D, 
28G, 33L, 34G, 35R, 36G, 
37P, 41V, 43D, 46C, 48R, 
68G, 70N, 71P, 79L, 82G, 

90G, 97N, 98G, 100H, 
101P, 147W, 156F, 160G, 
163P, 169G, 170F, 171D, 
173D, 174G, 182G, 185F, 
186P, 189P, 191D, 192W, 
196Y, 197P, 200P, 210G, 
211Y, 216F, 221S, 225G, 
232F, 246P, 256G, 261P, 

264G, 266W, 270Q, 286F, 
288G, 289D, 292G, 294P, 
297W, 308C, 321F, 324P, 
331W, 336F, 339P, 355P, 
374P, 402W, 405C, 409C, 
426P, 433R, 434Y, 437C, 
443V, 444F, 447C, 459H, 
471H, 489P, 494G, 505L 

and 518F 

 
 
Protein Stability 
 
For the three variants found in PNKP, prediction scenarios were run for all possible amino 
acid changes. Regression models were used to determine the consequences of variations 
for the protein stability (Table 7) using the changes in ddG values and translating them 
into a binary results. 
 



 
29 

 

 

Table 7. Results of the predictions made for each simulated residue variation. 

Protein Stability 
Variant 

Gln517 Arg433 T408del 

Decreased 
G, A, V, N, I, K, M, R, 
D, H, E, F, W, Y, S, T, 

C and P 

G, A, V, N, I, K, M, R, 
D, H, L,  E, F, W, Y, 

S, T, C and P 

G, A, V, N, I, K, 
M, R, D, H, L C, P 

and Q 

Increased L None E, F, W, Y, S and T 

 
Prediction of protein functionality 
 
Since the variant NM_007254.3: c.1294_1295dup found in patient AT-042 has not been 
previously described in literature, a prediction of protein functionality was performed. It 
was anticipated that around 55 amino acids will be lost corresponding to the protein’s 
kinase region. Therefore, the variant was predicted to be deleterious with a Tree vote: 
190|10. The number before the vertical bar (|) always represents deleterious predictions 
and the number after the vertical bar benign predictions (Schwarz et al., 2014).  
 
Discussion 

In this work, we clinically described five Costa Rican patients presenting signs of three 
different neuropathies. Through molecular analysis we found five variants associated 
with ataxia in the samples analyzed. 
 
Disease associated mutations in KCNC3 and PMPCA were identified in two patients, AT-
026 and AT-038. Variant p.Arg420His in KCNC3 is linked to Spinocerebellar ataxia type 
13 (SCA13), a rare sub-type of spinocerebellar ataxias (Michael F. Waters et al., 2006). 
This protein is required for normal survival of cerebellar neurons because it plays a role 
in the regulation of the frequency, shape and duration of action potentials in Purkinje 
cells (Bateman et al., 2021). While the known pathogenic variants in KCNC3 are 
associated with different phenotypes, data to date are too limited to make any genotype-
phenotype correlations (M.F. Waters, 2020). Nonetheless p.Arg420His is predicted to 
cause loss of channel activity; decreased in protein abundance and protein stability; 
reduced localization to the plasma membrane and impaired N-glycosylation (Zhao, Zhu, 
& Thornhill, 2013).  
 
Variant p.Ala377Thr, found in another patient (AT-042), is associated to Spinocerebellar 
ataxia, autosomal recessive 2, the first pure cerebellar ataxia to be described in which the 
inheritance is autosomal recessive (Rosenberg & Khemani, 2015). Jobling et al., 2015 
found that this particular variant leads to more abnormalities of mitochondrial function 
and impacts the maturation process of frataxin (FXN), the protein which is depleted in 
Friedreich ataxia (FRDA). 
 
For patient AT-004, it was possible to perform additional testing in FXN, thus lead to a 
diagnosis of Friedreich ataxia, the most common of the inherited ataxias (Delatycki, 
Williamson, & Forrest, 2000) and caused by a GAA repeat, generally in intron one. FXN 



 
30 

 

 

is a nuclear-encoded mitochondrial iron chaperone involved in iron-sulfur biogenesis and 
heme biosynthesis (MIM 229300). Different studies have shown that the size of the GAA 
repeat length in each allele is important in predicting the age of onset and some features 
of FRDA and that there is a far greater contribution from the smaller than larger allele to 
disease parameters and complications (Delatycki et al., 2000). 
 
Regarding PNKP, more than 40 individuals carrying PNKP gene mutations have been 
reported around the world so far. We identified both homozygous and compound 
heterozygous mutations in seven of the 28 patients whose samples were analyzed. This 
results agree with the previously variants linked to AOA4/CMT2B reported in Costa Rica 
by Leal et al., 2018, p.Gln517Ter and p.Thr408del, leading us to suggest the possibility 
of a founder effect to explain why only these variants, with the exception of AT-042, 
were found.  
 
Additionally to both p.Gln517Ter and p.Thr408del, a variant that has not been previously 
described was found, p.Arg433ProfsTer35. The pathogenicity of this sequence alterations 
could not be verified trough molecular approaches since additional family members were 
not available for segregation analysis.  
 
Nevertheless, this mutation could be causative for the disease phenotype due to the 
suspected production of truncated, partially or fully functional protein, thus contributing 
to the protein deficiency that is the distinguishing feature of many recessive genetic 
disorders (Nickless, Bailis, & You, 2017). With this in mind, different predictive tests 
were carried out using bioinformatics methods that provide information on the degree of 
conservation, stability and protein functionality. 
 
Multiple sequence alignments were analyzed to identify coevolving residues using a 
method that involves Mutual Information (MI), this was based on its ability to identify 
underlying relationships such as coevolution between pairs of residues (Moreno-Brid & 
Ruiz-Nápoles, 2009) (De Juan et al., 2013). In the four super families of proteins 
associated to AOA, functionally important residues already known to cause each of the 
existing types were inferred among the highly conserved positions, as well as potential 
candidate sites potential candidates that, due to their degree of conservation, could cause 
phenotypes similar to those currently described if they were to mutate. One of the 
candidate sites identified through MI is 433R, the same mutated residue that was 
identified in patient AT-042, this finding can serve as a real proof of concept observed in 
a patient.  
 
Although not all identified sites that are highly conserved are disease-causing, 
consistency between predictions and observed reality is detected and worth of following 
up in other types of studies with a larger cohort. A possible explanation to understand 
why some of the predicted conserved residues may not have been associated with disease 
yet is that they may be so pathogenic that the change becomes incompatible with life.  
 
The candidate sites identified in the super families represent positions of interest that 
would require validation through clinical reports or experimental data. Our results 
suggest that MI may be a novel approach for the identification of new disease candidate 
sites, while providing valuable insight into how underlying amino acid relationships may 
shape their role in the occurrence of different phenotypic manifestations. 
 



 
31 

 

 

Genetic variations, such as hydrogen bonding networks, conformational dynamics, 
protein activity and protein interaction networks, particularly at the level of functional 
assemblies, can have dramatic effects on protein stability and constitute one of the main 
molecular mechanisms underlying several mutation-induced diseases (Sanavia et al., 
2020). Although the protein stability prediction tool used in this study performed the tests 
presented by calculating the effects of amino acid substitution and not with early 
termination, the result was a loss of stability in most scenarios for all three variants found 
in PNKP.  
 
Since it has been demonstrated that protein stability is a major factor contributing to 
monogenic disease (Yue, Li, & Moult, 2005), these results can be used as a basis for 
inferring that an early termination mutation would have a similar effect to the ones 
reviewed, if the mutation has any effect on the stability at the ddG level, either increased 
or decreased, resulting in a truncated protein, that would likely cause the protein to be 
inelastic or denatured and the checkpoints responsible for detecting these defects would 
proceed to discard it. 
 
However, to assess the detrimental effect of the variants found, protein stability is 
necessary but not sufficient to infer protein function, since proteins are not necessarily 
optimized to maximize their stability (Chen et al., 2020). Therefore, it was considered 
crucial to understanding how the unknown will functionally impact important sites. 
Functionality prediction analysis for p.Arg433ProfsTer35 indicated that the resulting 
amino acid sequence change will cause the loss of more than 10% of the protein structure 
(kinase domain) leading to increased DNA damage with subsequent cell death due to the 
loss of the catalytic function that enables the protein to transfer phosphate molecules 
usually from ATP to other substrates (Bateman et al., 2021).  
 
Taking into account that the resulting prediction points to such severity of the new 
variant, a possible explanation for why the AT-042 phenotype is less severe than expected 
goes hand in hand with the proposal by Bermúdez-Guzmán, et al., 2020 of mutational 
survivorship bias. This suggests a higher tolerance in the kinase domain, which groups 
most of the deleterious variants described to date, as opposed to the phosphatase domain. 
It is supported by previous studies proposing that the phosphatase domain is functionally 
more important and necessary for DNA repair and therefore there are fewer variants 
described as they may cause individuals to be non-viable. This theory could provide an 
explanation as to why in the present case of an individual with compound heterozygous 
condition (AT-042) in which the variants reported in the kinase domain are deleterious 
and even one allele is completely unusable, the enzymatic capacity of the protein is only 
diminished, and this allows some level of functionality to be expressed.  
 
The p.Thr408del is a known variant, present in homozygous form for AT-010 and as a 
compound heterozygous for 1071 and AT-042, the T408 has a location within the kinase 
region so its deletion probably has consequences for protein conformation and function 
(Bras et al., 2015). The second variant found in 1071, p.Gln517Ter, causes the loss of 
those amino acids of the enzyme, that play a role in the stabilization of the protein, 
anchoring the kinase domain to the phosphatase domain (Leal et al., 2018). 
 
 Residues located from positions 402–521 of PNKP are coevolved and considerably 
conserved among the respective superfamily, this suggest that the C-terminal tail, as well 
as its interactions are critical for the protein’s function which in response to ionizing 



 
32 

 

 

radiation or oxidative damage catalyzes 5' phosphorylation of nucleic acids and also has 
an associated 3' phosphatase activity, predicting an important role in DNA repair after 
ionizing radiation or oxidative damage, multiple pathways involved in DNA-damage 
repair, including single-strand breaks (SSBs) and double-strand breaks (DSB) (Bras et 
al., 2015; Uhlen et al., 2015). 
 
In this way, the molecular involvement of PNKP in the different pathways mentioned 
above, might contribute to the disease phenotypes associated with the deletion 
(p.Thr408del), the premature termination (p.Gln517Ter), as well as the potential 
candidate (Arg433ProfsTer) since this defects can result in protein alterations leading to 
damaged DNA, mainly due to oxidative stress in the nervous system, with ensuing 
interference with transcription, reduced enzyme activity and, finally, cell death. 
 
In the case of the 19 patients in whom no mutations were found, through whole exome 
sequencing, our hypothesis is related to the sequencing method used, since there are 
regions and genetic variants that cannot be technically covered by whole exome 
sequencing method such as structural chromosomal aberrations, trinucleotide repeat 
expansion, epigenetic factors, variants in genes with corresponding pseudogenes or other 
highly homologous sequences, and noncoding regions including untranslated regions, 
introns, and intergenic regions. For this patients, further testing will be required to reach 
a diagnosis. 
 
In conclusion, this work lays the groundwork for molecular diagnosis of neuropathies 
and ataxias associated with this gene, as well as in other genes involved in peripheral 
neuropathies, and would generate phenotypic data related to ataxias in Costa Rica. To the 
best of our knowledge, this is the first time that homozygous p.Thr408del patients are 
reported in Costa Rica, and these findings along to those by Leal et al, (2018) suggest 
that mutations in PNKP are the most frequent cause of AOA in the country.  
 
Our investigation also provides an approach to prioritize mutations discovered in large-
scale sequencing projects which serves to validate these bioinformatics tools and 
recognize their predictive method in indicating sites of interest. 
 
 
Acknowledgments 
 
We would like to thank the patients and families for their participation in this study. We 
would also like to thank M.Sc. Melissa Vásquez Cerdas for her collaborations in regards 
to the samples located in the INISA, MBBS, PhD. Sanjay I. Bidichandani, CHF Claire 
Gordon Duncan Chair in Genetics, University of Oklahoma College of Medicine for 
performing the test for FRDA diagnosis, Lidia Benitez and Marcelo Castro-Alpízar for 
their valuable input while performing tests and writing this paper. 
 
 
References 
 
Altschul, S. F., Madden, T. L., Schäffer, A. A., Zhang, J., Zhang, Z., Miller, W., & 

Lipman, D. J. (1997). Gapped BLAST and PSI-BLAST:a new generation of protein 
database search programs. Nucleic Acids Research, 25(17), 3389–3402. 
https://doi.org/10.1093/nar/25.17.3389 



 
33 

 

 

Bateman, A., Martin, M. J., Orchard, S., Magrane, M., Agivetova, R., Ahmad, S., … 
Zhang, J. (2021). UniProt: The universal protein knowledgebase in 2021. Nucleic 
Acids Research, 49(D1), D480–D489. https://doi.org/10.1093/nar/gkaa1100 

Becherel, O. J., Sun, J., Yeo, A. J., Nayler, S., Fogel, B. L., Gao, F., … Lavin, M. F. 
(2015). A new model to study neurodegeneration in ataxia oculomotor apraxia type 
2. Human Molecular Genetics, 24(20), 5759–5774. 
https://doi.org/10.1093/hmg/ddv296 

Bermúdez-Guzmán, L., Jimenez-Huezo, G., Arguedas, A., & Leal, A. (2020). Mutational 
survivorship bias: The case of PNKP. PLoS ONE, 15(12 December), 1–25. 
https://doi.org/10.1371/journal.pone.0237682 

Bird, T. D. (2019). Hereditary Ataxia Overview. Retrieved from GeneReviewsTM 
website: https://www.ncbi.nlm.nih.gov/books/NBK1138/ 

Bras, J., Alonso, I., Barbot, C., Costa, M. M., Darwent, L., Orme, T., … Guerreiro, R. 
(2015). Mutations in PNKP cause recessive ataxia with oculomotor apraxia type 4. 
American Journal of Human Genetics, 96(3), 474–479. 
https://doi.org/10.1016/j.ajhg.2015.01.005 

Brenner, C. (2002). Hint, Fhit, and GalT: Function, structure, evolution, and mechanism 
of three branches of the histidine triad superfamily of nucleotide hydrolases and 
transferases. Biochemistry, 41(29), 9003–9014. https://doi.org/10.1021/bi025942q 

Brenner, C., Biecanowski, P., Pace, H. C., & Huebner, K. (1999). The histidine triad 
superfamily of nucleotide-binding proteins. Journal of Cellular Physiology, 181(2), 
179–187. https://doi.org/10.1002/(SICI)1097-4652(199911)181:2<179::AID-
JCP1>3.0.CO;2-8 

Chen, C. W., Lin, M. H., Liao, C. C., Chang, H. P., & Chu, Y. W. (2020). iStable 2.0: 
Predicting protein thermal stability changes by integrating various characteristic 
modules. Computational and Structural Biotechnology Journal, 18, 622–630. 
https://doi.org/10.1016/j.csbj.2020.02.021 

Choquet, K., Zurita-Rendón, O., La Piana, R., Yang, S., Dicaire, M. J., Boycott, K. M., … 
Tétreault, M. (2016). Autosomal recessive cerebellar ataxia caused by a 
homozygous mutation in PMPCA. Brain, 139(3), e19. 
https://doi.org/10.1093/brain/awv362 

Coutinho, P., & Barbot, C. (2002). Ataxia with Oculomotor Apraxia Type 1. 
GeneReviewsTM, 1–12. https://doi.org/NBK1456 [bookaccession] 

De Juan, D., Pazos, F., & Valencia, A. (2013). Emerging methods in protein co-evolution. 
Nature Reviews Genetics, 14(4), 249–261. https://doi.org/10.1038/nrg3414 

Delatycki, M. B., Williamson, R., & Forrest, S. M. (2000). Friedreich ataxia: An 
overview. Journal of Medical Genetics, 37(1), 1–8. 
https://doi.org/10.1136/jmg.37.1.1 

Engelman, J. A., Luo, J., & Cantley, L. C. (2006). The evolution of phosphatidylinositol 
3-kinases as regulators of growth and metabolism. Nature Reviews Genetics, 7(8), 
606–619. https://doi.org/10.1038/nrg1879 

Gatti, M., Magri, S., Nanetti, L., Sarto, E., Di Bella, D., Salsano, E., … Taroni, F. (2019). 
From congenital microcephaly to adult onset cerebellar ataxia: Distinct and 
overlapping phenotypes in patients with PNKP gene mutations. American Journal 
of Medical Genetics, Part A, 179(11), 2277–2283. 
https://doi.org/10.1002/ajmg.a.61339 

Greene, D., Richardson, S., & Turro, E. (2016). Phenotype Similarity Regression for 
Identifying the Genetic Determinants of Rare Diseases. American Journal of Human 
Genetics, 98(3), 490–499. https://doi.org/10.1016/j.ajhg.2016.01.008 

Guevara Coto, J., Schwartz, C. E., & Wang, L. (2014). Protein sector analysis for the 



 
34 

 

 

clustering of disease-associated mutations. BMC Genomics, 15(Suppl 11), S4. 
https://doi.org/10.1186/1471-2164-15-S11-S4 

Halabi, N., Rivoire, O., Leibler, S., & Ranganathan, R. (2009). Protein Sectors: 
Evolutionary Units of Three-Dimensional Structure. Cell, 138(4), 774–786. 
https://doi.org/10.1016/j.cell.2009.07.038 

Holmes, G. (1908). An attempt to classify cerebellar disease , with a note on Marie’s 
Hereditary Cerebellar Ataxia. Brain, 30(4), 545–567. 
https://doi.org/https://doi.org/10.1093/brain/30.4.545 

Howe, K. L., Achuthan, P., Allen, J., Allen, J., Alvarez-Jarreta, J., Ridwan Amode, M., … 
Flicek, P. (2021). Ensembl 2021. Nucleic Acids Research, 49(D1), D884–D891. 
https://doi.org/10.1093/nar/gkaa942 

Inlora, J., Sailani, M. R., Khodadadi, H., Teymurinezhad, A., Takahashi, S., Bernstein, J. 
A., … Snyder, M. P. (2017). Identification of a novel mutation in the APTX gene 
associated with ataxia-oculomotor apraxia. Cold Spring Harbor Molecular Case 
Studies, 3(6), 1–10. https://doi.org/10.1101/mcs.a002014 

Jayadev, S., & Bird, T. D. (2013). Hereditary ataxias: Overview. Genetics in Medicine, 
15(9), 673–683. https://doi.org/10.1038/gim.2013.28 

Jobling, R. K., Assoum, M., Gakh, O., Blaser, S., Raiman, J. A., Mignot, C., … Yoon, 
G. (2015). PMPCA mutations cause abnormal mitochondrial protein processing in 
patients with non-progressive cerebellar ataxia. Brain, 138(6), 1505–1517. 
https://doi.org/10.1093/brain/awv057 

Karczewski, K. J., Francioli, L. C., Tiao, G., Cummings, B. B., Alföldi, J., Wang, Q., … 
MacArthur, D. G. (2020). The mutational constraint spectrum quantified from 
variation in 141,456 humans. Nature, 581(7809), 434–443. 
https://doi.org/10.1038/s41586-020-2308-7 

Kent, W., Sugnet, C., Furey, T., Roskin, K., Pringle, T., Zahler, A., & Haussler, D. 
(2002). The human genome browser at UCSC. Retrieved March 10, 2018, from 
https://genome.ucsc.edu/cgi-
bin/hgTracks?db=hg38&lastVirtModeType=default&lastVirtModeExtraState=&vi
rtModeType=default&virtMode=0&nonVirtPosition=&position=chr9%3A329726
06-33001628&hgsid=679854365_a5kEusvN9FluLmsqGveaiqayHGGD 

Klockgether, T. (2010). Sporadic ataxia with adult onset: classification and diagnostic 
criteria. The Lancet Neurology, 9(1), 94–104. https://doi.org/10.1016/S1474-
4422(09)70305-9 

Köhler, S., Gargano, M., Matentzoglu, N., Carmody, L. C., Lewis-Smith, D., Vasilevsky, 
N. A., … Robinson, P. N. (2021). The human phenotype ontology in 2021. Nucleic 
Acids Research, 49(D1), D1207–D1217. https://doi.org/10.1093/nar/gkaa1043 

Köhler, S., Schulz, M. H., Krawitz, P., Bauer, S., Dölken, S., Ott, C. E., … Robinson, P. 
N. (2009). Clinical Diagnostics in Human Genetics with Semantic Similarity 
Searches in Ontologies. American Journal of Human Genetics, 85(4), 457–464. 
https://doi.org/10.1016/j.ajhg.2009.09.003 

Kumar, S., Stecher, G., Tamura, K., & Dudley, J. (2016). MEGA7: Molecular 
Evolutionary Genetics Analysis Version 7.0 for Bigger Datasets. Mol. Biol. Evol, 
33(7), 1870–1874. https://doi.org/10.1093/molbev/msw054 

Landrum, M. J., Lee, J. M., Benson, M., Brown, G. R., Chao, C., Chitipiralla, S., … 
Magglott, D. R. (2018). ClinVAr: improving access to variant interpretations and 
supporting evidence. Nucleic Acids Research, 46, D1062–D1067. 
https://doi.org/DOI: 10.1093/nar/gkx1153 

Leal, A., Bogantes-Ledezma, S., Ekici, A. B., Uebe, S., Thiel, C. T., Sticht, H., … Reis, 
A. (2018). The polynucleotide kinase 3’-phosphatase gene (PNKP) in Charcot-



 
35 

 

 

Marie-Tooth disease (CMT2B2) previously related to MED25. Neurogenetics. 
https://doi.org/https://doi.org/10.1007/s1048-018-0555-7 

Li, W., Jaroszewski, L., & Godzik, A. (2001). Clustering of highly homologous 
sequences to reduce the size of large protein databases. In BIOINFORMATICS 
APPLICATIONS NOTE (Vol. 17). Retrieved from http://bioinformatics. 

Manto, M. U. (2005). The wide spectrum of spinocerebellar ataxias (SCAs). Cerebellum, 
4(1), 2–6. https://doi.org/10.1080/14734220510007914 

Mariani, L. L., Rivaud-Péchoux, S., Charles, P., Ewenczyk, C., Meneret, A., Monga, B. 
B., … Anheim, M. (2017). Comparing ataxias with oculomotor apraxia: A 
multimodal study of AOA1, AOA2 and at focusing on video-oculography and 
alpha-fetoprotein. Scientific Reports, 7(1), 1–9. https://doi.org/10.1038/s41598-017-
15127-9 

McLaughlin, R. N., Poelwijk, F. J., Raman, A., Gosal, W. S., & Ranganathan, R. (2012). 
The spatial architecture of protein function and adaptation. Nature, 491(7422), 138–
142. https://doi.org/10.1038/nature11500 

Medicine;, M.-N. I. of G., & Johns Hopkins University (Baltimore, M. (n.d.). Online 
Mendelian Inheritance in Man, OMIM®. Retrieved March 10, 2018, from 
https://omim.org/ 

Morcos, F., Pagnani, A., Lunt, B., Bertolino, A., Marks, D. S., Sander, C., … Weigt, M. 
(2011). Direct-coupling analysis of residue coevolution captures native contacts 
across many protein families. Proceedings of the National Academy of Sciences of 
the United States of America, 108(49). https://doi.org/10.1073/pnas.1111471108 

Moreira, M. C., Klur, S., Watanabe, M., Németh, A. H., Le Ber, I., Moniz, J. C., … 
Koenig, M. (2004). Senataxin, the ortholog of a yeast RNA helicase, is mutant in 
ataxia-ocular apraxia 2. Nature Genetics, 36(3), 225–227. 
https://doi.org/10.1038/ng1303 

Moreira, M., & Koenig, M. (n.d.). Ataxia with Oculomotor Apraxia Type 2 (E. Adam 
MP, Ardinger HH, Pagon RA, et al., Ed.). Retrieved from GeneReviews® [Internet] 
website: https://www.ncbi.nlm.nih.gov/books/NBK1154/ 

Moreno-Brid, J. C., & Ruiz-Nápoles, P. (2009). La educación superior y el desarrollo 
económico en América Latina. Estudios y Perspectivas CEPAL, 106, 46. 

Nemeth, A. H., Bochukova, E., Dunne, E., Huson, S. M., Elston, J., Hannan, M. A., … 
Taylor, A. M. R. (2000). Autosomal recessive cerebellar ataxia with oculomotor 
apraxia (ataxia-telangiectasia-like syndrome) is linked to chromosome 9q34. 
American Journal of Human Genetics, 67(5), 1320–1326. 
https://doi.org/10.1016/S0002-9297(07)62962-0 

Nickless, A., Bailis, J. M., & You, Z. (2017). Control of gene expression through the 
nonsense-mediated RNA decay pathway. Cell and Bioscience, 7(1), 1–12. 
https://doi.org/10.1186/s13578-017-0153-7 

Pedroso, J. L., Rocha, C. R. R., Macedo-Souza, L. I., De Mario, V., Marques, W., 
Barsottini, O. G. P., … Kok, F. (2015). Clinical/scientific notes. Neurology: 
Genetics, 1(4), 1–4. https://doi.org/10.1212/NXG.0000000000000030 

Perlman, S. (2022). Hereditary Ataxia Overview 1 . Clinical Characteristics of Primary 
Hereditary Ataxia. Retrieved from https://www.ncbi.nlm.nih.gov/books/NBK1138/ 

Prasad, T. S. K., Goel, R., Kandasamy, K., Keerthikumar, S., Kumar, S., Mathivanan, 
S., … Kishore, A. (2009). Human Protein Reference Database. Retrieved March 12, 
2018, from http://hprd.org/index_html 

Pulst, S. M., Nechiporuk, A., Nechiporuk, T., Gispert, S., Chen, X. N., Lopes-Cendes, 
I., … Sahba, S. (1996). Moderate expansion of a normally biallelic trinucleotide 
repeat in spinooerebellar ataxia type. Nature Genetics, 14(3), 269–276. 



 
36 

 

 

https://doi.org/10.1038/ng1196-269 

Richard, P., Feng, S., & Manley, J. L. (2013a). A SUMO-dependent interaction between 
Senataxin and the exosome, disrupted in the neurodegenerative disease AOA2, 
targets the exosome to sites of transcription-induced DNA damage. Genes and 
Development, 27(20), 2227–2232. https://doi.org/10.1101/gad.224923.113 

Richard, P., Feng, S., & Manley, J. L. (2013b). A SUMO-dependent interaction between 
Senataxin and the exosome, disrupted in the neurodegenerative disease AOA2, 
targets the exosome to sites of transcription-induced DNA damage. Genes and 
Development, 27(20), 2227–2232. https://doi.org/10.1101/gad.224923.113 

Richards, S., Aziz, N., Bale, S., Bick, D., Das, S., Gastier-Foster, J., … Rehm, H. L. 
(2015). Standards and guidelines for the interpretation of sequence variants: A joint 
consensus recommendation of the American College of Medical Genetics and 
Genomics and the Association for Molecular Pathology. Genetics in Medicine, 
17(5), 405–424. https://doi.org/10.1038/gim.2015.30 

Rosenberg, R. N., & Khemani, P. (2015). The Inherited Ataxias. In Rosenberg’s 
Molecular and Genetic Basis of Neurological and Psychiatric Disease: Fifth Edition 
(Fifth Edit, pp. 811–832). https://doi.org/10.1016/B978-0-12-410529-4.00071-1 

Ruano, L., Melo, C., Silva, M. C., & Coutinho, P. (2014). The global epidemiology of 
hereditary ataxia and spastic paraplegia: A systematic review of prevalence studies. 
Neuroepidemiology, 42(3), 174–183. https://doi.org/10.1159/000358801 

Sanavia, T., Birolo, G., Montanucci, L., Turina, P., Capriotti, E., & Fariselli, P. (2020). 
Limitations and challenges in protein stability prediction upon genome variations: 
towards future applications in precision medicine. Computational and Structural 
Biotechnology Journal, 18, 1968–1979. https://doi.org/10.1016/j.csbj.2020.07.011 

Schiess, N., Zee, D. S., Siddiqui, K. A., Szolics, M., & El-Hattab, A. W. (2017). Novel 
PNKP mutation in siblings with ataxia-oculomotor apraxia type 4. Journal of 
Neurogenetics, 31(1–2), 23–25. https://doi.org/10.1080/01677063.2017.1322079 

Schmitz-Hübsch, T., Coudert, M., Bauer, P., Giunti, P., Globas, C., Baliko, L., … 
Klockgether, T. (2008). Spinocerebellar ataxia types 1, 2, 3, and 6: Disease severity 
and nonataxia symptoms. Neurology, 71(13), 982–989. 
https://doi.org/10.1212/01.wnl.0000325057.33666.72 

Schöls, Ludger; Bauer, Peter; Schmidt, Thorsten; Schulte Thorsten; Riess, O. (2004). 
Autosomal dominant cerebellar ataxias: clinical features, genetics and pathogenesis. 
Neurology, 3(2), 8. https://doi.org/10.4324/9780429456916-3 

Schwarz, J. M., Cooper, D. N., Schuelke, M., & Seelow, D. (2014). Mutationtaster2: 
Mutation prediction for the deep-sequencing age. Nature Methods, 11(4), 361–362. 
https://doi.org/10.1038/nmeth.2890 

Seo, G. H., Kim, T., Choi, I. H., Park, J. young, Lee, J., Kim, S., … Lee, B. H. (2020). 
Diagnostic yield and clinical utility of whole exome sequencing using an automated 
variant prioritization system, EVIDENCE. Clinical Genetics, 98(6), 562–570. 
https://doi.org/10.1111/cge.13848 

Simonetti, F. L., Teppa, E., Chernomoretz, A., Nielsen, M., & Marino Buslje, C. (2013). 
MISTIC: Mutual information server to infer coevolution. Nucleic Acids Research, 
41(Web Server issue), 8–14. https://doi.org/10.1093/nar/gkt427 

Tassan, N. Al, Khalil, D., Shinwari, J., Sharif, L. Al, Bavi, P., Abduljaleel, Z., … 
Bohlega, S. (2012). A Missense mutation in PIK3R5 gene in a family with ataxia 
and oculomotor apraxia. Human Mutation, 33(2), 351–354. 
https://doi.org/10.1002/humu.21650 

Thompson, J. D., Gibson, T. J., & Higgins, D. G. (2002). Multiple Sequence Alignment 
Using ClustalW and ClustalX. Current Protocols in Bioinformatics, 1–22. 



 
37 

 

 

https://doi.org/10.1002/0471250953.bi0203s00 

Uhlen, M., Fagerberg, L., Hallstrom, B. M., Lindskog, C., Oksvold, P., Mardinoglu, 
A., … Ponten, F. (2015). Tissue-based map of the human proteome. Science, 
347(6220), 1260419–1260419. https://doi.org/10.1126/science.1260419 

van Minkelen, R., Guitart, M., Escofet, C., Yoon, G., Elfferich, P., Bolman, G. M., … 
van den Ouweland, A. M. W. (2015). Complete APTX deletion in a patient with 
ataxia with oculomotor apraxia type 1. BMC Medical Genetics, 16(1), 1–4. 
https://doi.org/10.1186/s12881-015-0213-y 

Waters, M. F., Fee, D., Figueroa, K. P., Nolte, D., Müller, U., Advincula, J., … Pulst, S. 
M. (2005). An autosomal dominant ataxia maps to 19q13: Allelic heterogeneity of 
SCA13 or novel locus? Neurology, 65(7), 1111–1113. 
https://doi.org/10.1212/01.wnl.0000177490.05162.41 

Waters, Michael F. (2020). Spinocerebellar Ataxia Type 13 - GeneReviews® - NCBI 
Bookshelf. Retrieved June 19, 2022, from 
https://www.ncbi.nlm.nih.gov/books/NBK1225/ 

Waters, Michael F., Minassian, N. A., Stevanin, G., Figueroa, K. P., Bannister, J. P. A., 
Nolte, D., … Pulst, S. M. (2006). Mutations in voltage-gated potassium channel 
KCNC3 cause degenerative and developmental central nervous system phenotypes. 
Nature Genetics, 38(4), 447–451. https://doi.org/10.1038/ng1758 

Yue, P., Li, Z., & Moult, J. (2005). Loss of protein structure stability as a major causative 
factor in monogenic disease. Journal of Molecular Biology, 353(2), 459–473. 
https://doi.org/10.1016/j.jmb.2005.08.020 

Zhao, J., Zhu, J., & Thornhill, W. B. (2013). Spinocerebellar ataxia-13 Kv3.3 potassium 
channels: Arginine-to-histidine mutations affect both functional and protein 
expression on the cell surface. Biochemical Journal, 454(2), 259–265. 
https://doi.org/10.1042/BJ20130034 

Zühlke, C., Gehlken, U., Hellenbroich, Y., Schwinger, E., & Bürk, K. (2003). 
Phenotypical variability of expanded alleles in the TATA-binding protein gene: 
Reduced penetrance in SCA17? Journal of Neurology, 250(2), 161–163. 
https://doi.org/10.1007/s00415-003-0958-7 

 

 

 

 

 

 



 
38 

 

 

Conclusiones 

En el mundo se han identificado cuatro genes causantes de los diferentes tipos de AOA 

identificados hasta la fecha, en este estudio se identificaron en PNKP las variantes 

p.Gln517Ter y p.Thr408del que están asociadas a AOA4 en condición tanto homocigota 

como heterocigota compuesta. Estas variantes habían sido caracterizadas en estudios 

previos como causantes de AOA4/CMT2B en la población de Costa Rica. 

Además de las variantes mencionadas previamente, se identificó una variante no descrita 

en la literatura (Arg433Ter) y que los análisis bioinformáticos de predicción de 

estabilidad y funcionalidad proteica señalan como una posible causante de la enfermedad. 

Sin embargo, más estudios son necesarios para confirmar la hipótesis que relaciona el 

fenotipo observado con esta nueva variante.  

Las mutaciones identificadas en PNKP -deleciones y responsables de terminaciones 

tempranas- pueden dar lugar a alteraciones proteicas que conducen a ineficiencia de 

reparación del ADN dañado. De hecho, debido a estas mutaciones, , se desestabiliza la 

estructura proteica por debilitamiento del anclaje del dominio quinasa al dominio 

fosfatasa, se reduce de la actividad enzimática (fosforilación 5' de ácidos nucleicos y